Re: [time-nuts] Help w/integration problem
Never mind, I think I see what's wrong... you can't integrate the dBc/Hz
values directly. You have to turn them back into linear ratios, do the
interval sum, and then, if you want dBc coming out, take 10*log10(sum).
-- john, KE5FX
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Behalf Of John Miles
Sent: Monday, January 02, 2006 12:00 AM
To: Discussion of precise time and frequency measurement
Subject: [time-nuts] Help w/integration problem
Does anyone have a piece of C (BASIC, whatever) code that turns
an array of
dBc/Hz values into integrated RMS noise?
I'm trying to use a simple rectangular integrator to divide a log-log plot
into bins:
for (i=L_column; i U_column-1; i++)
{
sum += ((value[i] - ((value[i] - value[i+1]) / 2.0)) *
(frequency[i+1] - frequency[i]));
}
This just takes the midpoint dBc/Hz value between successive columns of a
phase-noise plot, multiplies it by the frequency step between the
columns in
question, and sums the result for all columns in the range of interest.
The output of this process, when I feed a typical noise graph with values
around -110 dBc/Hz to it, with frequency values at the lower and upper
limits of 1000 and 1 Hz, is around -1E+6. What I'd *like* is a value
corresponding to the -63 dBc value cited on pages 7 and 8 in
this Zarlink
app note:
http://assets.zarlink.com/CA/Phase_Noise_and_Jitter_Article.pdf
In this note, the author shows a noise curve similar to the ones
I'm working
with, and magically pulls -63 dBc out of the ether with no explanation of
the integration process that obtained it. (What does it mean, in the
author's words, to take the area under a phase-noise curve, anyway?
What's the bottom dBc/Hz value?)
Being from the instant-gratification generation, I really don't want (and
won't understand) a calculus lecture. I want the 5 lines of code that do
the integration. :-) This is for the next release of my freeware GPIB
noise-measurement app, so your karma will be integrated along
with the noise
if you're able to help!
-- john, KE5FX
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
From: John Miles [EMAIL PROTECTED]
Subject: Re: [time-nuts] Help w/integration problem
Date: Mon, 2 Jan 2006 00:49:19 -0800
Message-ID: [EMAIL PROTECTED]
Never mind, I think I see what's wrong... you can't integrate the dBc/Hz
values directly. You have to turn them back into linear ratios, do the
interval sum, and then, if you want dBc coming out, take 10*log10(sum).
You are almost there... you need to square the linears sum, which is quickly
done...
sum = 0
sum = sum + pow(10,value[1]/10)
sum = sum + pow(10,value[2]/10)
...
sum = sum + pow(10,value[n]/10)
rms = sqrt(sum)
dBc = 10 * log10(sum)
Normally you would use pow(10,value[1]/20) etc. to get the amplitudes back, but
RMS is about summing the power and that is the amplitude square as you recall.
Hmm... I'm less a math-freak this morning than usual. A good morning it is
anyway.
Cheers,
Magnus
-- john, KE5FX
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Behalf Of John Miles
Sent: Monday, January 02, 2006 12:00 AM
To: Discussion of precise time and frequency measurement
Subject: [time-nuts] Help w/integration problem
Does anyone have a piece of C (BASIC, whatever) code that turns
an array of
dBc/Hz values into integrated RMS noise?
I'm trying to use a simple rectangular integrator to divide a log-log plot
into bins:
for (i=L_column; i U_column-1; i++)
{
sum += ((value[i] - ((value[i] - value[i+1]) / 2.0)) *
(frequency[i+1] - frequency[i]));
}
This just takes the midpoint dBc/Hz value between successive columns of a
phase-noise plot, multiplies it by the frequency step between the
columns in
question, and sums the result for all columns in the range of interest.
The output of this process, when I feed a typical noise graph with values
around -110 dBc/Hz to it, with frequency values at the lower and upper
limits of 1000 and 1 Hz, is around -1E+6. What I'd *like* is a value
corresponding to the -63 dBc value cited on pages 7 and 8 in
this Zarlink
app note:
http://assets.zarlink.com/CA/Phase_Noise_and_Jitter_Article.pdf
In this note, the author shows a noise curve similar to the ones
I'm working
with, and magically pulls -63 dBc out of the ether with no explanation of
the integration process that obtained it. (What does it mean, in the
author's words, to take the area under a phase-noise curve, anyway?
What's the bottom dBc/Hz value?)
Being from the instant-gratification generation, I really don't want (and
won't understand) a calculus lecture. I want the 5 lines of code that do
the integration. :-) This is for the next release of my freeware GPIB
noise-measurement app, so your karma will be integrated along
with the noise
if you're able to help!
-- john, KE5FX
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
Thanks; yes, I've got the sqrt() part already, from both my original source who requested the feature, and the Zarlink app note. Naturally, the two sources don't agree. Equation 13 (and others) in the Maxim app note at http://pdfserv.maxim-ic.com/en/an/AN3359.pdf says, in effect: RMS = sqrt(sum * 2) On page 7 of the Zarlink app note, the x2 factor is left outside the radical sign: RMS = sqrt(sum) * 2 Unlike the question of whether to interpolate the column midpoints in dBc space or linear spectral-density space, the position of that x2 term makes a big difference in the final result. Any insights into who's got THAT one right? -- john, KE5FX -Original Message- From: Magnus Danielson [mailto:[EMAIL PROTECTED] Sent: Monday, January 02, 2006 1:28 AM To: time-nuts@febo.com; [EMAIL PROTECTED] Subject: Re: [time-nuts] Help w/integration problem From: John Miles [EMAIL PROTECTED] Subject: Re: [time-nuts] Help w/integration problem Date: Mon, 2 Jan 2006 00:49:19 -0800 Message-ID: [EMAIL PROTECTED] Never mind, I think I see what's wrong... you can't integrate the dBc/Hz values directly. You have to turn them back into linear ratios, do the interval sum, and then, if you want dBc coming out, take 10*log10(sum). You are almost there... you need to square the linears sum, which is quickly done... sum = 0 sum = sum + pow(10,value[1]/10) sum = sum + pow(10,value[2]/10) ... sum = sum + pow(10,value[n]/10) rms = sqrt(sum) dBc = 10 * log10(sum) Normally you would use pow(10,value[1]/20) etc. to get the amplitudes back, but RMS is about summing the power and that is the amplitude square as you recall. Hmm... I'm less a math-freak this morning than usual. A good morning it is anyway. ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
From: John Miles [EMAIL PROTECTED] Subject: Re: [time-nuts] Help w/integration problem Date: Mon, 2 Jan 2006 01:44:33 -0800 Message-ID: [EMAIL PROTECTED] Thanks; yes, I've got the sqrt() part already, from both my original source who requested the feature, and the Zarlink app note. I didn't bother to look at the Zarlink app note. Until now. Naturally, the two sources don't agree. Equation 13 (and others) in the Maxim app note at http://pdfserv.maxim-ic.com/en/an/AN3359.pdf says, in effect: RMS = sqrt(sum * 2) Yes, the magic happends between (11) and (12). The integration is 0 to infinity and not -infinity to infinity, since we already know it mirrors arround 0. Mind you that these are twice the power, not twice the amplitude. The energy at fc-f will have the same energy and be coherent to the energy at fc+f, so these energies add up perfectly. There is a special-case when you can't argue like this, but we can look the other way here and pick out the real reference literature when we need to. On page 7 of the Zarlink app note, the x2 factor is left outside the radical sign: RMS = sqrt(sum) * 2 Looks like sloppy work to me compared to the Maxim paper, which gives motivation to the formulas. Unlike the question of whether to interpolate the column midpoints in dBc space or linear spectral-density space, the position of that x2 term makes a big difference in the final result. Any insights into who's got THAT one right? I hope you've got some insight on that. I could dig deeper into the issue if you are not quite satisfied. I have better references than the two PDFs you mentioned. The whole single-sides/double-side spectra issue is a bit confusing and painstaking at first, I know. Cheers, Magnus -- john, KE5FX -Original Message- From: Magnus Danielson [mailto:[EMAIL PROTECTED] Sent: Monday, January 02, 2006 1:28 AM To: time-nuts@febo.com; [EMAIL PROTECTED] Subject: Re: [time-nuts] Help w/integration problem From: John Miles [EMAIL PROTECTED] Subject: Re: [time-nuts] Help w/integration problem Date: Mon, 2 Jan 2006 00:49:19 -0800 Message-ID: [EMAIL PROTECTED] Never mind, I think I see what's wrong... you can't integrate the dBc/Hz values directly. You have to turn them back into linear ratios, do the interval sum, and then, if you want dBc coming out, take 10*log10(sum). You are almost there... you need to square the linears sum, which is quickly done... sum = 0 sum = sum + pow(10,value[1]/10) sum = sum + pow(10,value[2]/10) ... sum = sum + pow(10,value[n]/10) rms = sqrt(sum) dBc = 10 * log10(sum) Normally you would use pow(10,value[1]/20) etc. to get the amplitudes back, but RMS is about summing the power and that is the amplitude square as you recall. Hmm... I'm less a math-freak this morning than usual. A good morning it is anyway. ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help - Hope?
Sorry for top posting, but I'm shifting the original topic, and apologies for a long semi-off topic post. I found the original post great and respect both of you guys very much. Tonight, (perhaps some spirits, in the holiday spirit, are influencing me, but) I thought about this discussion in a general case. I learned from smart people and was stimulated to know that I could hang out with people smarter than me. I find this conversation a great example of the best use of the internet. John has done much work in many realms, and has shared a good amount of his very useful creations and experience with the rest of the world on the internet. Magnus, I know less about but he, seems to be one of the people I would go to for guidance on any number of subjects. This medium has brought these two, geographically widely separate (I assume), people together on this conversation that will probably benefit many with John's software. I applaud you both for your willingness to share your knowledge and many of the products of that knowledge. But tonight, I'm thinking about the state of the world, in general. This is a private, and limited mailing list, but its members seem to reflect what I see on public newsgroups. The majority of contributions seem to be by older people. John is young by most of these standards. Popular folklore says that the internet is populated with young people. So my question: Are the younger people no longer attracted to the basic questions of science and engineering, or am I just missing the messages from young people for some reason? I know this mailing list is not typical, but I don't see younger people anywhere I go to share knowledge. Wish they were there. On Mon, 02 Jan 2006 11:42:53 +0100 (CET), Magnus Danielson [EMAIL PROTECTED] wrote: From: John Miles [EMAIL PROTECTED] Subject: Re: [time-nuts] Help w/integration problem Date: Mon, 2 Jan 2006 01:44:33 -0800 Message-ID: [EMAIL PROTECTED] Thanks; yes, I've got the sqrt() part already, from both my original source who requested the feature, and the Zarlink app note. I didn't bother to look at the Zarlink app note. Until now. Naturally, the two sources don't agree. Equation 13 (and others) in the Maxim app note at http://pdfserv.maxim-ic.com/en/an/AN3359.pdf says, in effect: RMS = sqrt(sum * 2) Yes, the magic happends between (11) and (12). The integration is 0 to infinity and not -infinity to infinity, since we already know it mirrors arround 0. Mind you that these are twice the power, not twice the amplitude. The energy at fc-f will have the same energy and be coherent to the energy at fc+f, so these energies add up perfectly. There is a special-case when you can't argue like this, but we can look the other way here and pick out the real reference literature when we need to. On page 7 of the Zarlink app note, the x2 factor is left outside the radical sign: RMS = sqrt(sum) * 2 Looks like sloppy work to me compared to the Maxim paper, which gives motivation to the formulas. Unlike the question of whether to interpolate the column midpoints in dBc space or linear spectral-density space, the position of that x2 term makes a big difference in the final result. Any insights into who's got THAT one right? I hope you've got some insight on that. I could dig deeper into the issue if you are not quite satisfied. I have better references than the two PDFs you mentioned. The whole single-sides/double-side spectra issue is a bit confusing and painstaking at first, I know. Cheers, Magnus -- john, KE5FX [snip first message] ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
Hi Mike --
I can't speak for John, but I'd sure find that program useful.
73,
John
Mike Feher said the following on 01/02/2006 08:10 AM:
John -
I wrote a program about 20 years ago to calculate the total integrated noise
power from the individual power spectrum density points. It is in GW Basic
and I still use it almost daily. Let me know if you would like it. 73 - Mike
Mike B. Feher, N4FS
89 Arnold Blvd.
Howell, NJ, 07731
732-886-5960
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of John Miles
Sent: Monday, January 02, 2006 3:00 AM
To: Discussion of precise time and frequency measurement
Subject: [time-nuts] Help w/integration problem
Does anyone have a piece of C (BASIC, whatever) code that turns an array of
dBc/Hz values into integrated RMS noise?
I'm trying to use a simple rectangular integrator to divide a log-log plot
into bins:
for (i=L_column; i U_column-1; i++)
{
sum += ((value[i] - ((value[i] - value[i+1]) / 2.0)) *
(frequency[i+1] - frequency[i]));
}
This just takes the midpoint dBc/Hz value between successive columns of a
phase-noise plot, multiplies it by the frequency step between the columns in
question, and sums the result for all columns in the range of interest.
The output of this process, when I feed a typical noise graph with values
around -110 dBc/Hz to it, with frequency values at the lower and upper
limits of 1000 and 1 Hz, is around -1E+6. What I'd *like* is a value
corresponding to the -63 dBc value cited on pages 7 and 8 in this Zarlink
app note:
http://assets.zarlink.com/CA/Phase_Noise_and_Jitter_Article.pdf
In this note, the author shows a noise curve similar to the ones I'm working
with, and magically pulls -63 dBc out of the ether with no explanation of
the integration process that obtained it. (What does it mean, in the
author's words, to take the area under a phase-noise curve, anyway?
What's the bottom dBc/Hz value?)
Being from the instant-gratification generation, I really don't want (and
won't understand) a calculus lecture. I want the 5 lines of code that do
the integration. :-) This is for the next release of my freeware GPIB
noise-measurement app, so your karma will be integrated along with the noise
if you're able to help!
-- john, KE5FX
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help - Hope?
In message [EMAIL PROTECTED], Magnus Danielson writes: But to answer your question, younger people is still attracted and there is still plenty of people having the right mind for these things around. A major difference for these younger people is that the technology of today is reverse engineering resistant. There is practically nothing to learn today by taking things apart: you can't see how they work. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 [EMAIL PROTECTED] | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
John -
Do you also need GW Basic? I'll FTP the programs and send the URL shortly.
73 - Mike
Mike B. Feher, N4FS
89 Arnold Blvd.
Howell, NJ, 07731
732-886-5960
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of John Ackermann N8UR
Sent: Monday, January 02, 2006 9:09 AM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Help w/integration problem
Hi Mike --
I can't speak for John, but I'd sure find that program useful.
73,
John
Mike Feher said the following on 01/02/2006 08:10 AM:
John -
I wrote a program about 20 years ago to calculate the total integrated
noise
power from the individual power spectrum density points. It is in GW Basic
and I still use it almost daily. Let me know if you would like it. 73 -
Mike
Mike B. Feher, N4FS
89 Arnold Blvd.
Howell, NJ, 07731
732-886-5960
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of John Miles
Sent: Monday, January 02, 2006 3:00 AM
To: Discussion of precise time and frequency measurement
Subject: [time-nuts] Help w/integration problem
Does anyone have a piece of C (BASIC, whatever) code that turns an array
of
dBc/Hz values into integrated RMS noise?
I'm trying to use a simple rectangular integrator to divide a log-log plot
into bins:
for (i=L_column; i U_column-1; i++)
{
sum += ((value[i] - ((value[i] - value[i+1]) / 2.0)) *
(frequency[i+1] - frequency[i]));
}
This just takes the midpoint dBc/Hz value between successive columns of a
phase-noise plot, multiplies it by the frequency step between the columns
in
question, and sums the result for all columns in the range of interest.
The output of this process, when I feed a typical noise graph with values
around -110 dBc/Hz to it, with frequency values at the lower and upper
limits of 1000 and 1 Hz, is around -1E+6. What I'd *like* is a value
corresponding to the -63 dBc value cited on pages 7 and 8 in this
Zarlink
app note:
http://assets.zarlink.com/CA/Phase_Noise_and_Jitter_Article.pdf
In this note, the author shows a noise curve similar to the ones I'm
working
with, and magically pulls -63 dBc out of the ether with no explanation of
the integration process that obtained it. (What does it mean, in the
author's words, to take the area under a phase-noise curve, anyway?
What's the bottom dBc/Hz value?)
Being from the instant-gratification generation, I really don't want (and
won't understand) a calculus lecture. I want the 5 lines of code that do
the integration. :-) This is for the next release of my freeware GPIB
noise-measurement app, so your karma will be integrated along with the
noise
if you're able to help!
-- john, KE5FX
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
Hi Mike --
Actually, I'm most likely to convert it to Perl or Python or something
to run on my Linux boxes. I'll be happy to make the port available to
anyone who wants it (no guarantees when it'll be done, though).
73,
John
Mike Feher said the following on 01/02/2006 09:51 AM:
John -
Do you also need GW Basic? I'll FTP the programs and send the URL shortly.
73 - Mike
Mike B. Feher, N4FS
89 Arnold Blvd.
Howell, NJ, 07731
732-886-5960
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of John Ackermann N8UR
Sent: Monday, January 02, 2006 9:09 AM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Help w/integration problem
Hi Mike --
I can't speak for John, but I'd sure find that program useful.
73,
John
Mike Feher said the following on 01/02/2006 08:10 AM:
John -
I wrote a program about 20 years ago to calculate the total integrated
noise
power from the individual power spectrum density points. It is in GW Basic
and I still use it almost daily. Let me know if you would like it. 73 -
Mike
Mike B. Feher, N4FS
89 Arnold Blvd.
Howell, NJ, 07731
732-886-5960
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of John Miles
Sent: Monday, January 02, 2006 3:00 AM
To: Discussion of precise time and frequency measurement
Subject: [time-nuts] Help w/integration problem
Does anyone have a piece of C (BASIC, whatever) code that turns an array
of
dBc/Hz values into integrated RMS noise?
I'm trying to use a simple rectangular integrator to divide a log-log plot
into bins:
for (i=L_column; i U_column-1; i++)
{
sum += ((value[i] - ((value[i] - value[i+1]) / 2.0)) *
(frequency[i+1] - frequency[i]));
}
This just takes the midpoint dBc/Hz value between successive columns of a
phase-noise plot, multiplies it by the frequency step between the columns
in
question, and sums the result for all columns in the range of interest.
The output of this process, when I feed a typical noise graph with values
around -110 dBc/Hz to it, with frequency values at the lower and upper
limits of 1000 and 1 Hz, is around -1E+6. What I'd *like* is a value
corresponding to the -63 dBc value cited on pages 7 and 8 in this
Zarlink
app note:
http://assets.zarlink.com/CA/Phase_Noise_and_Jitter_Article.pdf
In this note, the author shows a noise curve similar to the ones I'm
working
with, and magically pulls -63 dBc out of the ether with no explanation of
the integration process that obtained it. (What does it mean, in the
author's words, to take the area under a phase-noise curve, anyway?
What's the bottom dBc/Hz value?)
Being from the instant-gratification generation, I really don't want (and
won't understand) a calculus lecture. I want the 5 lines of code that do
the integration. :-) This is for the next release of my freeware GPIB
noise-measurement app, so your karma will be integrated along with the
noise
if you're able to help!
-- john, KE5FX
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
I figured that is what you would do. As I recall in this program I doubled
the answer to obtain the total double sided noise power. You will see that
in the program, towards the end, and can eliminate the doubling if you only
want the single sided total power over the selected integration limits. Here
is the URL:
http://[EMAIL PROTECTED]/DOS/PHASE1.BAS
Mike B. Feher, N4FS
89 Arnold Blvd.
Howell, NJ, 07731
732-886-5960
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of John Ackermann N8UR
Sent: Monday, January 02, 2006 10:02 AM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Help w/integration problem
Hi Mike --
Actually, I'm most likely to convert it to Perl or Python or something
to run on my Linux boxes. I'll be happy to make the port available to
anyone who wants it (no guarantees when it'll be done, though).
73,
John
Mike Feher said the following on 01/02/2006 09:51 AM:
John -
Do you also need GW Basic? I'll FTP the programs and send the URL shortly.
73 - Mike
Mike B. Feher, N4FS
89 Arnold Blvd.
Howell, NJ, 07731
732-886-5960
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of John Ackermann N8UR
Sent: Monday, January 02, 2006 9:09 AM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Help w/integration problem
Hi Mike --
I can't speak for John, but I'd sure find that program useful.
73,
John
Mike Feher said the following on 01/02/2006 08:10 AM:
John -
I wrote a program about 20 years ago to calculate the total integrated
noise
power from the individual power spectrum density points. It is in GW Basic
and I still use it almost daily. Let me know if you would like it. 73 -
Mike
Mike B. Feher, N4FS
89 Arnold Blvd.
Howell, NJ, 07731
732-886-5960
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of John Miles
Sent: Monday, January 02, 2006 3:00 AM
To: Discussion of precise time and frequency measurement
Subject: [time-nuts] Help w/integration problem
Does anyone have a piece of C (BASIC, whatever) code that turns an array
of
dBc/Hz values into integrated RMS noise?
I'm trying to use a simple rectangular integrator to divide a log-log plot
into bins:
for (i=L_column; i U_column-1; i++)
{
sum += ((value[i] - ((value[i] - value[i+1]) / 2.0)) *
(frequency[i+1] - frequency[i]));
}
This just takes the midpoint dBc/Hz value between successive columns of a
phase-noise plot, multiplies it by the frequency step between the columns
in
question, and sums the result for all columns in the range of interest.
The output of this process, when I feed a typical noise graph with values
around -110 dBc/Hz to it, with frequency values at the lower and upper
limits of 1000 and 1 Hz, is around -1E+6. What I'd *like* is a value
corresponding to the -63 dBc value cited on pages 7 and 8 in this
Zarlink
app note:
http://assets.zarlink.com/CA/Phase_Noise_and_Jitter_Article.pdf
In this note, the author shows a noise curve similar to the ones I'm
working
with, and magically pulls -63 dBc out of the ether with no explanation of
the integration process that obtained it. (What does it mean, in the
author's words, to take the area under a phase-noise curve, anyway?
What's the bottom dBc/Hz value?)
Being from the instant-gratification generation, I really don't want (and
won't understand) a calculus lecture. I want the 5 lines of code that do
the integration. :-) This is for the next release of my freeware GPIB
noise-measurement app, so your karma will be integrated along with the
noise
if you're able to help!
-- john, KE5FX
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
___
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help - Hope?
¡Felices fiestas y un próspero año nuevo! Kind of a bummer of a thread to start a new year. Thought I'd demonstrate that anything remains possible by contributing a message that doesn't have anything to do with leap seconds :-) Poul-Henning Kamp makes a good point: A major difference for these younger people is that the technology of today is reverse engineering resistant. There is practically nothing to learn today by taking things apart: you can't see how they work ... a good point as far as commercial electronics is concerned, for instance. Budding home experimenters and hobbiests do now have a significantly higher hurdle to clear if they seek to understand the operation of their equipment. The boxes are blacker than they've ever been. ...on the other hand, no black box can be truly opaque and remain operational. Things, of course, come in an ever wider variety of flavors - many of which rely rather closely (most certainly including timing issues) on fundamental physics that cannot ultimately be hidden from view. Seeing how they work doesn't have to be restricted to disassembly and visual inspection. The operation of inspection may include various probes and scopes. Disassembly may include physical deconstruction, sure, but may also include software techniques to understand algorithms - or any other sequence of operations intended to understand the interrelationship of subsystems. In fact, several interesting reverse engineering techniques rely on completely non-invasive techniques. Codes are cracked by monitoring the power consumption of smart cards - devices with no moving parts. Perhaps internet mailing lists are seeing a lull in subscriptions from new devotees. (Or perhaps the young'ns simply can't get a word in edgewise :-) I'm unaware of any decrease in enrollment in technical disciplines in the academic community. Some departments are growing and some are shrinking as the balance shifts from hardware to software to bioware - but the overall level of interest is surely growing. Start a list focused on biological clocks and see how much interest you get. In fact, one suspects that the natural lifecycle of a mailing list involves a burst of interest (and subscriptions) in the beginning followed by a long tail. Mailing lists in general are most certainly mortal. On the issue of the world's breadth of technical insight and enthusiasm, there are no reasons to fret that weren't outlined in Zen and the Art of Motorcycle Maintenance. Rob Seaman National Optical Astronomy Observatory ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
[time-nuts] Looking for data
Greetings, A friend at work has the following recordings available over the leap second, and will post them if there's interest. DCF77 (.de) WWV, WWVB (.us) MSF (.uk) BPM (.ch) This station went to carrier at leap second. oops. He's interested in the following: CHU (.ca) WWVH(.us) JJY (.jp) TDF (.fr) HBC (swiss) Warner ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Looking for data
I have CHU at http://www.febo.com/time-freq/leapsecond-2005 (about halfway down the page). John Warner Losh said the following on 01/02/2006 12:29 PM: Greetings, A friend at work has the following recordings available over the leap second, and will post them if there's interest. DCF77 (.de) WWV, WWVB (.us) MSF (.uk) BPM (.ch) This station went to carrier at leap second. oops. He's interested in the following: CHU (.ca) WWVH(.us) JJY (.jp) TDF (.fr) HBC (swiss) Warner ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
Hi, Mike -- That would indeed be interesting to see... but your file looks like a compiled (or at least tokenized) binary. Is there an interpreter for Win2K/XP that will let me list your program, or some other way to get a plaintext listing? BTW, I don't know if you saw the note I posted to the hp_agilent list or not, but I've got PN.EXE, SSM.EXE, and 7470.EXE running on the 8568A now. As soon as I get this noise-integration feature nailed down, I'll build a new release that should work on your 8566A. Will be in touch offline. -- john, KE5FX -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Mike Feher Sent: Monday, January 02, 2006 7:12 AM To: 'Discussion of precise time and frequency measurement' Subject: Re: [time-nuts] Help w/integration problem I figured that is what you would do. As I recall in this program I doubled the answer to obtain the total double sided noise power. You will see that in the program, towards the end, and can eliminate the doubling if you only want the single sided total power over the selected integration limits. Here is the URL: http://[EMAIL PROTECTED]/DOS/PHASE1.BAS ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
Yes, the magic happends between (11) and (12). The integration is 0 to infinity and not -infinity to infinity, since we already know it mirrors arround 0. Mind you that these are twice the power, not twice the amplitude. The energy at fc-f will have the same energy and be coherent to the energy at fc+f, so these energies add up perfectly. There is a special-case when you can't argue like this, but we can look the other way here and pick out the real reference literature when we need to. Thanks; you're right, given the integration limits in the Maxim note, their way makes more sense. On page 7 of the Zarlink app note, the x2 factor is left outside the radical sign: RMS = sqrt(sum) * 2 Looks like sloppy work to me compared to the Maxim paper, which gives motivation to the formulas. Agreed. I'll leave the *2 operation inside the radicand. Much appreciate the help! -- john, KE5FX ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
From: John Miles [EMAIL PROTECTED] Subject: Re: [time-nuts] Help w/integration problem Date: Mon, 2 Jan 2006 11:42:33 -0800 Message-ID: [EMAIL PROTECTED] Yes, the magic happends between (11) and (12). The integration is 0 to infinity and not -infinity to infinity, since we already know it mirrors arround 0. Mind you that these are twice the power, not twice the amplitude. The energy at fc-f will have the same energy and be coherent to the energy at fc+f, so these energies add up perfectly. There is a special-case when you can't argue like this, but we can look the other way here and pick out the real reference literature when we need to. Thanks; you're right, given the integration limits in the Maxim note, their way makes more sense. Indeed. It took some time to get sure, but once I was sure it was obvious. On page 7 of the Zarlink app note, the x2 factor is left outside the radical sign: RMS = sqrt(sum) * 2 Looks like sloppy work to me compared to the Maxim paper, which gives motivation to the formulas. Agreed. I'll leave the *2 operation inside the radicand. Much appreciate the help! Anytime! Also, if you look at the Maxim paper, it rather looks like a graphical error not to extend the squareroot sign all the way. The logical place to put the 2 if not included in the square-root is actually before as customary. Cheers, Magnus ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help - Hope?
In message [EMAIL PROTECTED], Rob Seaman writes: =A1Felices fiestas y un pr=F3spero a=F1o nuevo! ...on the other hand, no black box can be truly opaque and remain operational. Things, [...] cannot ultimately be hidden from view. Seeing how they work doesn't have to be restricted to disassembly and visual inspection. That is true for a person of sufficient dedication, just like swimming the Channel is possible for a suitably dedicated person. But what we're after here is not people psyked and trained to cross the Channel, we're talking about how to interest kids in going into cold water, dressed in practically nothing and spend time to learn to swim. The main difference between technology and swimming however, is that the wares we peddle seldom offer glimpses of female anatomy in even skimpier clothing. Kids interest in how the world works peaks at age 7-12, and if you have to teach them disassembly before you can satisfy their curiosity, the technological world is in deep trouble indeed. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 [EMAIL PROTECTED] | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Looking for data
In message [EMAIL PROTECTED], Warner Losh writes: Greetings, A friend at work has the following recordings available over the leap second, and will post them if there's interest. DCF77 (.de) WWV, WWVB (.us) MSF (.uk) BPM (.ch) This station went to carrier at leap second. oops. Isn't this China ? .cn ? He's interested in the following: TDF (.fr) HBC (swiss) I have 400 GB of VLF/LF data, sampled a 5 MSPS * 12 bits directly from my loop antenna. It should contain not only HBC and TDF, but also DCF77, Rugby, LORAN-C but also anything else up to around 300 kHz (low-pass filter in my antenna) -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 [EMAIL PROTECTED] | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help - Hope?
I think we're seeing the technology shift to a different level of abstraction, that's all. If the operating principles of a system built from components cannot be understood in structural terms (i.e., from disassembly), then your definition of component is what's insufficient. You just need to move up a level and try again. Even in this rarefied company, few of us truly work from first principles. A veteran RFIC designer may well have forgotten Maxwell's equations and all of their implications. I'm not familiar with the details of RFIC modelling, design, and fabrication, but I understand how the end product is applied at the circuit level. A kid playing with a WiFi card and a Pringles can doesn't know the first thing about how his Wifi card works, but he will, after some experimenting, understand how to use it to talk to his neighbor's access point. To the chip designer, a component is probably a subcircuit model that exists only in software. To me, a component is the resulting chip, with pins you can solder stuff to. To the kid, the component is the monolithic WiFi card. There is little to be gained by assigning relative levels of merit to different abstraction levels, or assuming that society is doomed because people rarely work their way down the abstraction hierarchy without a compelling need. I'm happiest when I'm able to work all the way down the lowest level of abstraction I can actually put my hands on, but that doesn't carry much weight in a global sense. Witness the trouble I run into when I actually _need_ the first principles that I ran away from when I dropped out of college. :-) -- john, KE5FX -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Poul-Henning Kamp Sent: Monday, January 02, 2006 6:46 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Help - Hope? In message [EMAIL PROTECTED], Magnus Danielson writes: But to answer your question, younger people is still attracted and there is still plenty of people having the right mind for these things around. A major difference for these younger people is that the technology of today is reverse engineering resistant. There is practically nothing to learn today by taking things apart: you can't see how they work. ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Jupiter GPS receiver
Can someone point me to the source where theknown fault of the Jupiter listed below is documented?? I have a bunch of these engines and would like to see this documentation. Tnx! 73, Mike, N1JEZ A closed mouth gathers no feet - Original Message - From: Bjorn Gabrielsson [EMAIL PROTECTED] To: Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Sunday, January 01, 2006 4:57 PM Subject: Re: [time-nuts] Jupiter GPS receiver Geoff [EMAIL PROTECTED] writes: Didier Juges wrote: I am planning to let the GPS receiver run a while longer to see if it corrects itself at some point, and if it does not, I'll reboot it. Hello Didier, The Jupiter GPS receiver (if used in NMEA and not binary mode) has a known fault, that it can be 1 or 2 seconds delta to UTC, this is independent of the leap-second situation. Hi Geoff! This afternoon, :-( ,I got an opportunity to run my jupiter in binary mode again. It had been running in NMEA mode, when switched to binary it was 2 seconds late in binary mode. But had bit set saying that 'time mark not valid' After some 15 minutes it ran ok again, still do. First strange behaviour I have seen in binary mode. Regret missing the leap second, but I would not be surprised if my jupiter would have shown problems. Anybody running a jupiter in binary mode over the event? -- Björn. ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Jupiter GPS receiver
Hello Mike, Can someone point me to the source where theknown fault of the Jupiter listed below is documented?? Google has a cached copy of a discussion that I participated in, back in Feb 2004. Try this URL, I think it should work: http://tinyurl.com/c6bfn It contains the statement from Navman (the current manufacturers) acknowledging the NMEA issue. Mike, I have no details about what make or model of the Jupiter this NMEA issue applies to. It may be generic from the first Rockwell unit to Navman, all I raise is a question mark regarding the NMEA timestamps with the Jupiter. If you have a Jupiter and can generate an NMEA logfile over a few days Mike, I wrote a simple utility that can race through an NMEA logfile and find any sequence errors in the time (like the Jupiter can make). If you wanted to test yours I would be happy to send you the wee app (DOS but runs under windows). Regards, Geoff. ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
Magnus - So far I do not have any problems running any of my basic programs for myself. The problem comes when I try to share like this time. All the best in 2006 - Mike Mike B. Feher 89 Arnold Blvd. Howell, NJ, 07731 732-886-5960 -Original Message- From: Magnus Danielson [mailto:[EMAIL PROTECTED] Sent: Monday, January 02, 2006 8:42 PM To: time-nuts@febo.com; [EMAIL PROTECTED] Subject: Re: [time-nuts] Help w/integration problem From: Mike Feher [EMAIL PROTECTED] Subject: Re: [time-nuts] Help w/integration problem Date: Mon, 2 Jan 2006 20:26:54 -0500 Message-ID: [EMAIL PROTECTED] Mike, The URL below is the source code in Basic for the program that I wrote over 20 years ago. Since I think you both intend to change it to another language the listing should suffice. I have so many great programs in Basic, some that go back over 30 years, but, not too many machines support DOS easily any more. I still use these programs however. I guess it is easier than me to learn a new programming language. 73 - Mike http://www.eozinc.com/DOS/Phase1.pdf I'm quite sure that it is possible to find more or less suitable BASIC implementations lying around. For UNIX style OSes there is for instance the Bywater BASIC (bwBASIC) https://sourceforge.net/projects/bwbasic/ and Yeat Another BASIC (yaBASIC) http://www.yabasic.de/. There is a fair amount of HPBASIC programs around which would be good if one could run. The main problem should be porting the GPIB interface. Anyway, for normal BASIC programs, there is certainly options around. BTW, I have a Euroboard (160x100mm) computer with a i8052AHBASIC V1.1 chip. Just toss power on it, a serial port to the propper pins and I have a little BASIC engine. This little board also has the EPROM socket, 8255s and other goodies, so it is a real little powerhouse for its time. ;O) The manuals for the BASIC (very detailed) is online. ;O) Cheers, Magnus ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
Thanks, Mike. I can work with that. 73, John Mike Feher said the following on 01/02/2006 08:26 PM: The URL below is the source code in Basic for the program that I wrote over 20 years ago. Since I think you both intend to change it to another language the listing should suffice. I have so many great programs in Basic, some that go back over 30 years, but, not too many machines support DOS easily any more. I still use these programs however. I guess it is easier than me to learn a new programming language. 73 - Mike http://www.eozinc.com/DOS/Phase1.pdf Mike B. Feher 89 Arnold Blvd. Howell, NJ, 07731 732-886-5960 -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of John Miles Sent: Monday, January 02, 2006 2:30 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Help w/integration problem Hi, Mike -- That would indeed be interesting to see... but your file looks like a compiled (or at least tokenized) binary. Is there an interpreter for Win2K/XP that will let me list your program, or some other way to get a plaintext listing? BTW, I don't know if you saw the note I posted to the hp_agilent list or not, but I've got PN.EXE, SSM.EXE, and 7470.EXE running on the 8568A now. As soon as I get this noise-integration feature nailed down, I'll build a new release that should work on your 8566A. Will be in touch offline. -- john, KE5FX -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Mike Feher Sent: Monday, January 02, 2006 7:12 AM To: 'Discussion of precise time and frequency measurement' Subject: Re: [time-nuts] Help w/integration problem I figured that is what you would do. As I recall in this program I doubled the answer to obtain the total double sided noise power. You will see that in the program, towards the end, and can eliminate the doubling if you only want the single sided total power over the selected integration limits. Here is the URL: http://[EMAIL PROTECTED]/DOS/PHASE1.BAS ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Web page updated
At 12:21 PM 1/1/2006 -0500, John Ackermann N8UR wrote: I've wrapped a web page around my data, with photos of the Spectracom non-event :-) and some other info. http://www.febo.com/time-freq/leapsecond-2005. Any chance of getting a few more minutes (say leap-10 to leap+10) of WWVB? I'm still looking to see when the leap second warning bits were turned off. -- newell ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Web page updated
I've put the full 2-hour run of WWVB at http://www.febo.com/time-freq/leapsecond-2005/WWVB-full.mp3 (7MB) and http://www.febo.com/time-freq/leapsecond-2005/WWVB-ful.wav (115MB). Feel free to grab 'em (please try the MP3 first and only get the WAV if the MP3 won't do the job). John Scott Newell said the following on 01/02/2006 09:24 PM: At 12:21 PM 1/1/2006 -0500, John Ackermann N8UR wrote: I've wrapped a web page around my data, with photos of the Spectracom non-event :-) and some other info. http://www.febo.com/time-freq/leapsecond-2005. Any chance of getting a few more minutes (say leap-10 to leap+10) of WWVB? I'm still looking to see when the leap second warning bits were turned off. ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Web page updated
John Ackermann N8UR said the following on 01/02/2006 09:43 PM: I've put the full 2-hour run of WWVB at http://www.febo.com/time-freq/leapsecond-2005/WWVB-full.mp3 (7MB) and http://www.febo.com/time-freq/leapsecond-2005/WWVB-ful.wav (115MB). Sorry about the typo -- the second URL should be spelled just like the first, but for the extension. John ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Web page updated
At 09:43 PM 1/2/2006 -0500, John Ackermann N8UR wrote: Feel free to grab 'em (please try the MP3 first and only get the WAV if the MP3 won't do the job). Your MP3 worked just fine, even with the dirt simple AM demodulator I'm futzing with. Got the same results as with the demodulated datastream I recorded: the leap second warning bit doesn't turn off until the end of minute 00:02. thanks! -- newell ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
http://pdfserv.maxim-ic.com/en/an/AN3359.pdf is the Maxim note I was referring to. This HP app note, describing an advanced noise-measurement program for the 8568A, is excellent as well: http://www.speakeasy.net/~jmiles1/an270-2.pdf With a graph acquired from the junker 8568A I bought for GPIB work, I'm getting essentially-identical results to what they show in figure 4. Makes me a lot more confident in the math, as well as the condition of my $450 8568A. -- john, KE5FX Could someone please give me a reference for the Maxim application note that has been mentioned several times? I would like to see it, as when I wrote my program it was based on my own knowledge, and I want to make sure we are in agreement. Although, in the many years that I have used my program I have not found it to be faulty. Pretty straight forward stuff really. But, there have been times in the past when I thought that, and made mistakes. - Thanks - Mike Mike B. Feher 89 Arnold Blvd. Howell, NJ, 07731 732-886-5960 ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help w/integration problem
John - Thanks, good basic stuff. Right now my biggest concern is the amount of acceptable phase jitter for a given data rate for higher order modulations. I know that the rule of thumb is not to exceed 10% of the Euclidian distance between phase points in the constellation, but, how do you obtain the RMS phase jitter for a given symbol rate? I have seen literature that states you have to integrate all the way to Rs, but for the higher orders above BPSK, I think you only need to integrate to Rs/2. Naturally you still have to add 3 dB to the number prior to calculating the phase jitter in degrees due to the double sideband. I have pretty much convinced myself that going to Rs/2 is the upper bound of the integral. The question is, what is the lower bound? I have seen numbers as low as 1%, which seem absurd to me. I think even 10 % would probably not degrade BER for a given Eb/No to make a difference. Any ideas appreciated. - Thanks - Mike Mike B. Feher 89 Arnold Blvd. Howell, NJ, 07731 732-886-5960 -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of John Miles Sent: Monday, January 02, 2006 11:14 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Help w/integration problem http://pdfserv.maxim-ic.com/en/an/AN3359.pdf is the Maxim note I was referring to. This HP app note, describing an advanced noise-measurement program for the 8568A, is excellent as well: http://www.speakeasy.net/~jmiles1/an270-2.pdf With a graph acquired from the junker 8568A I bought for GPIB work, I'm getting essentially-identical results to what they show in figure 4. Makes me a lot more confident in the math, as well as the condition of my $450 8568A. -- john, KE5FX Could someone please give me a reference for the Maxim application note that has been mentioned several times? I would like to see it, as when I wrote my program it was based on my own knowledge, and I want to make sure we are in agreement. Although, in the many years that I have used my program I have not found it to be faulty. Pretty straight forward stuff really. But, there have been times in the past when I thought that, and made mistakes. - Thanks - Mike Mike B. Feher 89 Arnold Blvd. Howell, NJ, 07731 732-886-5960 ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help - Hope?
As a data point, I visited MIT last November and headed for selected spots, some last seen in 1960. The Edgerton Center on the 4th floor of Building 4 had a class in session doing things with 555 timers and LEDs. The instructor told me that they have resources for 12 students. They get 25 applications, from grad students to local high school students. Lessee, there are about 4000 MIT undergrads, and about 15 are interested in a hands-on lab. Maybe that's to be expected, but I remember how disappointed I was when the EE department tore out the rotating equipment lab and replaced it with courses in vector math in 1955. I switched to Mechanical Engineering because they hadn't gone completely abstract. OTOH, engineers live to create stuff with other people's money. The building blocks keep changing, but the urge to build is still there. I build computers with motherboards and power supplies and cases, etc. I build a time lab with boxes purchased from eBay. The thing is, we have lost the 7-12 group, the Boy Electricians, the Gilbert chemistry sets and the magic of radio. TV promised to be an exceptional teaching tool, but selfish people with an unending greed turned it into a behavioral modification tool to create consumers. Kids learn early to concentrate on consumption and forget about how the world works. The people with the most influence on kids don't want consumers that know how to think, especially not creatively. I can't do anything about it, although I did donate to the Edgerton Center, so I play with time and wait to see if Limits to Growth was right about the population collapse in 2020. Best wishes for the new year, but don't blame me if it keeps getting worse. Bill Hawkins ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help - Hope?
Good points Bill. Heck, if I was doing today what I was doing in my early teens I would be considered a terrorist. I go to LL quite a lot, and a lot of the newer hires there are MIT grads. At least most of the ones hired at LL do care, and somewhat understand hardware. Still, until they get at least 10 or 15 years under their belt they get confused between the esoteric mathematical gyrations and reality. - Mike Mike B. Feher 89 Arnold Blvd. Howell, NJ, 07731 732-886-5960 -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Bill Hawkins Sent: Monday, January 02, 2006 11:40 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Help - Hope? As a data point, I visited MIT last November and headed for selected spots, some last seen in 1960. The Edgerton Center on the 4th floor of Building 4 had a class in session doing things with 555 timers and LEDs. The instructor told me that they have resources for 12 students. They get 25 applications, from grad students to local high school students. Lessee, there are about 4000 MIT undergrads, and about 15 are interested in a hands-on lab. Maybe that's to be expected, but I remember how disappointed I was when the EE department tore out the rotating equipment lab and replaced it with courses in vector math in 1955. I switched to Mechanical Engineering because they hadn't gone completely abstract. OTOH, engineers live to create stuff with other people's money. The building blocks keep changing, but the urge to build is still there. I build computers with motherboards and power supplies and cases, etc. I build a time lab with boxes purchased from eBay. The thing is, we have lost the 7-12 group, the Boy Electricians, the Gilbert chemistry sets and the magic of radio. TV promised to be an exceptional teaching tool, but selfish people with an unending greed turned it into a behavioral modification tool to create consumers. Kids learn early to concentrate on consumption and forget about how the world works. The people with the most influence on kids don't want consumers that know how to think, especially not creatively. I can't do anything about it, although I did donate to the Edgerton Center, so I play with time and wait to see if Limits to Growth was right about the population collapse in 2020. Best wishes for the new year, but don't blame me if it keeps getting worse. Bill Hawkins ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] Help - Hope?
John Miles says: I think we're seeing the technology shift to a different level of abstraction Yes - this is certainly true of software, for instance. Our team attended JavaOne this year - along with 15,000 rabid (and much younger) technophiles. Object oriented programming replaces procedural programming replaces assembler coding replaces machine code - in significantly under one career length. As an undergraduate, I programmed a 6502 KIM to do productive work (plot time series photometry via an A/D connected to a photomultiplier) in machine code via its hex pad. Now you can generate OO code direct from UML (or so they claim - have yet to see it demonstrated practically). (Algorithms remain algorithms, however.) And no - physics remains physics. We're still building telescopes out of big shiny mirrors using optical principles well known to Fresnel and Fraunhofer. I'm reading the history of the first Atlantic telegraph cable. Great story full of details like Kelvin's invention of the precision galvanometer - virtually identical to the torsion devices whose mirrors I learned to read as an undergraduate. It may well be that TI or HP or Fluke will sell you a totally digital handheld gizmo with greater sensitivity (and features), but you still have to know as much about electrical circuits to use the new gizmos as you did the old gizmos. Meanwhile, it is apparently the case that today's cable laying ships still use cable handling techniques perfected during the travails of the first transatlantic cable venture 150 years ago. Some things change. Some things stay the same. What fundamentally remains the same is the reality underlying all our technology. Won't belabor the question of layering UTC on Earth orientation via mean solar time. Focus instead on the works of atomic clocks (or related gizmos like masers or whatever comes next). The levels of abstraction may be compressed to hide the details of intervening layers of complexity, but the two parts that will always remain are the user interface (itself an interesting reflection of human factors), and at the other end, the basic physics of whatever phenomena. I suspect I'm not alone on this list in volunteering as a local science fair judge. I focus on the middle school physical science projects as providing the most opportunity for encouraging a future career choice. Ignore the scoring rubric. The general award rules provide for first/second/third prizes with ties for second and third place, so the goal is simply to identify and rank the top five projects. It can be difficult (to put it mildly) to infer the mental state of most of the participants, but there are always a few that stand out (even after discounting the projects resulting from Science Olympiad, etc). Really, all that matters at that age is a sense of creativity. Actually, I weight any evidence of true curiosity and, well, fun even higher. These are rare (especially under the crushing weight of standards), but every year reveals new kids finding new ways of looking at familiar territory. I invariably leave more optimistic than when I arrived. Bottom line is that a scientific world view is likely no more prevalent now than it was in the middle ages. But it is likely no less prevalent, either. This is the world of Burning Man, the Long Now, and public key cryptography. My neighbor is an AF pilot whose son is rebuilding a Corvette from the ground up. These are not folks full of technological angst. We just happen to be in the natural pause between the first Moon landings and our inevitable (albeit politicized) return. We'll miss this time of relative quiet when it's gone. Suggesting that the love of technological pursuits is dying out is kind of like those 19th century ruminations that all of scientific knowledge was well in hand, or that guy who figured out several decades ago that all possible songs had already been written. Not too worried at this end. Rob Seaman National Optical Astronomy Observatory ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
