Driver to drive?

[Michael A. Terrell]

radams2000@gmail.com wrote:

Long ago I learned a painful lesson about video frequency accuracy. I assumed that getting a crystal within 100ppm of tbe ideal frequency should be just fine. But when we tried to send that signal through a satellite, it wouldn't go because the frequency accuracy was too poor.

I think the issue is that VCXO's are often used for clock recovery and they can't be pulled very far. Or maybe they just rely on absolute accuracy, and when things drift between the source and the sync, you just drop or repeat a frame. Not sure.

NTSC color calls for +/- 10 Hz for the 3,579,545 burst frequency.

 

[josephkk]

On Sat, 09 Feb 2013 01:08:47 -0800, Jamie M <jmorken@shaw.ca> wrote:

Hi,

I have an application (home brewing Kombucha) that requires a small
heater to increase the temperature for the fermentation to work
properly. Does anyone have a link to a cheap readily available
120VAC 40watt heater that can work in ambient heating the air with no
fan? I am using a 40watt incandescent lightbulb right now, but it
isn't ideal as the Kombucha is supposed to ferment with low light, and
tinfoil on the bulb causes too much overheating.

cheers,
Jamie

Lookup heat tape.

?-)

 

[josephkk]

On Mon, 11 Feb 2013 09:44:51 -0500, Spehro Pefhany
<speffSNIP@interlogDOTyou.knowwhat> wrote:

On Sun, 10 Feb 2013 19:45:28 -0800, John Larkin
jjlarkin@highNOTlandTHIStechnologyPART.com> wrote:


This 10 ps project may not be possible with parts on boards. The application
likely doesn't have enough volume for a custom IC, or even for wire-bonding up a
hybrid.

A general question- what kinds of circuits would benefit from a
wire-bonded circuit? I'm thinking high speed, low power, maybe low
noise. Probably size isn't much of an advantage these days.

Presumably availability of bare dice in small quantities is a major
issue?

(just general interest- I was thinking of learning how to do prototype
wirebonding just for the halibut).

Holy sharkskin suit, i was thinking along similar lines. With some fancy
Rogers materials for the "substrate" it may be cost effective for some
high performance applications.

?-)

 

[josephkk]

On Sun, 10 Feb 2013 13:27:47 +0100, Gerhard Hoffmann <dk4xp@arcor.de>
wrote:

Am 10.02.2013 01:34, schrieb John Larkin:
On Sun, 10 Feb 2013 00:24:58 +0100, Gerhard Hoffmann <dk4xp@arcor.de> wrote:

was meant as a email pointer to a colleague until we meet on Monday
Thunderbird is !x%$'!!! aargh!

That seems to be a hazard of using the same program for mail and newsgroups.

Why is the button for "create new" (verfassen in my German version) on
the top left and "answer newsgroup", forward, Archive on the right side
under the list of articles... I want my Agent back. But I won't return
to Windows for it.

I run Agent in wine. Works pretty well without any fuss to install and
get it running. Help doesn't work properly though.

BTW I'll play this weekend with your ramp generator (the one with the
AD8009 bootstrap). I have a new Altium Designer installation and
choose it as an exercise.

What's the application? Fast ramps are fun; fast ramps at 10-12 bit linearity
are even more fun.

This one is just to fathom how far I can go; I did something similar
with a switched current source in a time stretcher. We talked about that
last year.
It was for measuring the time of flight of just a few photons
(interpolation between clock edges)


http://www.google.de/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CDIQFjAA&url=http%3A%2F%2Fwww.ptb.de%2Femrp%2F1393.html%3F%26no_cache%3D1%26cid%3D2274%26did%3D2819%26sechash%3D40d6db49&ei=XYYXUZGEOsOm4gTrs4DoDg&usg=AFQjCNEAr10OW8j-gp15VWZdLIdkJ6C0FA&sig2=KO_mO3OuJETmIxdZAofhSQ&bvm=bv.42080656,d.Yms


page 33


The inductance of the ramp cap can be an issue if the discharge switch turns off
too fast. Sometimes I use a couple of caps in parallel, or slow down the turnoff
drive somehow. 1 ns speeds are interesting, but 100 ps is sort of a Chuck Yeager
style "wall in the sky."

This time I'd like to avoid the dual slope conversion and go
directly ramp to digital with a LTC2165. Avoids at least the
second switching, no droop in the slow part of the cycle,
no comparator bias current etc. And instead of a few hundred
stretch cycles per second I can get 20Meg at least.


Altium - Camtastic - pdf - offset film from nearby print shop -
empty board (self etched & drilled) in a long afternoon.

At some point I got tired of ferric chloride stains on my person. I just
breadboard on copperclad, or order a board and do something else until it
arrives.

I use Ammoniumpersulfat and presensitized 0.5mm FR4 made by Bungard.
6 mil traces and gaps are easy with offset film.
Strictly single sided plus GND layer, stamp-sized and dense.

With the Altium snippet feature I can recycle the circuit diagram and
the layout as a macro for the real design.
I now have quite a lot of pre-characterized stamps that I can use to
click a board together (regulators, VCXOs, filters, frequency
multipliers, phase comparators...)

regards, Gerhard

 

[josephkk]

On Sat, 09 Feb 2013 17:54:58 -0700, Jim Thompson
<To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:

On Sat, 09 Feb 2013 19:53:46 -0500, Spehro Pefhany
speffSNIP@interlogDOTyou.knowwhat> wrote:

On Sat, 09 Feb 2013 16:35:51 -0800, the renowned John Larkin
jjlarkin@highNOTlandTHIStechnologyPART.com> wrote:

On Sun, 10 Feb 2013 01:10:46 +0100, Gerhard Hoffmann <dk4xp@arcor.de> wrote:

Am 09.02.2013 21:51, schrieb John Larkin:
,
http://www.hittite.com/content/documents/data_sheet/hmc841lc4b.pdf



The clock-to-output prop delay is 10 ps typ. That's mind boggling. Light travels
across the width of the chip in more time than that.

And the rise/fall time to the decision point would be 6 ps of this.
That means, the "intent" to produce a different output level must travel
in 4 ps.

If you fudge the threshold definitions a little, you can claim negative prop
delay.

Resublimated thiotimoline semiconductors?


Best regards,
Spehro Pefhany

Naaaah! Just resublimated BS.

...Jim Thompson

I suspect that this one got past you.

You must not be a Dr. A fan.

?-)

 

[Jasen Betts]

On 2013-02-13, josephkk <joseph_barrett@sbcglobal.net> wrote:

I run Agent in wine. Works pretty well without any fuss to install and
get it running. Help doesn't work properly though.

hve you installed "xchm" or is help broken in lyspice+wine too?

--
⚂⚃ 100% natural

 

[dave]

On 02/12/2013 06:50 PM, josephkk wrote:

On Sat, 09 Feb 2013 01:08:47 -0800, Jamie M <jmorken@shaw.ca> wrote:

Hi,

I have an application (home brewing Kombucha) that requires a small
heater to increase the temperature for the fermentation to work
properly. Does anyone have a link to a cheap readily available
120VAC 40watt heater that can work in ambient heating the air with no
fan? I am using a 40watt incandescent lightbulb right now, but it
isn't ideal as the Kombucha is supposed to ferment with low light, and
tinfoil on the bulb causes too much overheating.

cheers,
Jamie

Lookup heat tape.

?-)

Hot plate and a light dimmer?

 

[dcaster@krl.org]

On Feb 12, 9:50 pm, josephkk <joseph_barr...@sbcglobal.net> wrote:

Lookup heat tape.

?-)

Just be sure you are not getting heat tape designed to keep things
from freezing. Some of them have high resistance at temperatures a
bit above freezing.

Dan

 

[josephkk]

On 13 Feb 2013 10:57:17 GMT, Jasen Betts <jasen@xnet.co.nz> wrote:

On 2013-02-13, josephkk <joseph_barrett@sbcglobal.net> wrote:

I run Agent in wine. Works pretty well without any fuss to install and
get it running. Help doesn't work properly though.

hve you installed "xchm" or is help broken in lyspice+wine too?

I have tried installing xchm and both are still broken. It is very
possible i did something wrong though.

?-)

 

[Jasen Betts]

On 2013-02-14, josephkk <joseph_barrett@sbcglobal.net> wrote:

On 13 Feb 2013 10:57:17 GMT, Jasen Betts <jasen@xnet.co.nz> wrote:

On 2013-02-13, josephkk <joseph_barrett@sbcglobal.net> wrote:

I run Agent in wine. Works pretty well without any fuss to install and
get it running. Help doesn't work properly though.

hve you installed "xchm" or is help broken in lyspice+wine too?

I have tried installing xchm and both are still broken. It is very
possible i did something wrong though.

?-)

from your descriptions it sounds kind of messed up.

you could try a fresh install by temporarily moving your ~/.wine
directory and reinstalling ltspice, if help and works in the new
install it's probablly something messed up in your wine install.

also test www with help -> about -> linear-tech-website


--
It's called a dead hammer because it's got shot in the head

 

[rickman]

On 2/12/2013 7:02 AM, radams2000@gmail.com wrote:

Long ago I learned a painful lesson about video frequency accuracy. I assumed that getting a crystal within 100ppm of tbe ideal frequency should be just fine. But when we tried to send that signal through a satellite, it wouldn't go because the frequency accuracy was too poor.

I think the issue is that VCXO's are often used for clock recovery and they can't be pulled very far. Or maybe they just rely on absolute accuracy, and when things drift between the source and the sync, you just drop or repeat a frame. Not sure.

I worked on a board for satcomms many years ago. We had a choice of
clocking an ADC at a fixed rate and using a few bits for "bit stuffing"
where either data was sent or a flag indicated no data, or we could have
used a VCXO which could be pulled some ą100 ppm. The software guy
didn't have any confidence in controlling the VCXO so we used the bit
stuffing. It simply meant we needed an extra 4 bits per frame.

--

Rick

 

[rickman]

On 2/12/2013 12:25 PM, Michael A. Terrell wrote:

radams2000@gmail.com wrote:

Long ago I learned a painful lesson about video frequency accuracy. I assumed that getting a crystal within 100ppm of tbe ideal frequency should be just fine. But when we tried to send that signal through a satellite, it wouldn't go because the frequency accuracy was too poor.

I think the issue is that VCXO's are often used for clock recovery and they can't be pulled very far. Or maybe they just rely on absolute accuracy, and when things drift between the source and the sync, you just drop or repeat a frame. Not sure.


NTSC color calls for +/- 10 Hz for the 3,579,545 burst frequency.

Hmmm... ą3 ppm. Did they have much luck finding crystals that would
provide this? I expect there is more than just a crystal in this
circuit. I seem to recall that the color burst *is* a reference that
lets the local TV oscillator sync up on every frame. Maybe the camera
has a ą3 ppm oscillator in it, but I seriously doubt TVs do.

--

Rick

 

[Michael A. Terrell]

rickman wrote:

On 2/12/2013 12:25 PM, Michael A. Terrell wrote:

radams2000@gmail.com wrote:

Long ago I learned a painful lesson about video frequency accuracy. I assumed that getting a crystal within 100ppm of tbe ideal frequency should be just fine. But when we tried to send that signal through a satellite, it wouldn't go because the frequency accuracy was too poor.

I think the issue is that VCXO's are often used for clock recovery and they can't be pulled very far. Or maybe they just rely on absolute accuracy, and when things drift between the source and the sync, you just drop or repeat a frame. Not sure.


NTSC color calls for +/- 10 Hz for the 3,579,545 burst frequency.

Hmmm... ą3 ppm. Did they have much luck finding crystals that would
provide this? I expect there is more than just a crystal in this
circuit. I seem to recall that the color burst *is* a reference that
lets the local TV oscillator sync up on every frame. Maybe the camera
has a ą3 ppm oscillator in it, but I seriously doubt TVs do.

Apparently you've never looked at the schematic of a color TV or
monitor or the RS-170A standard.

"Color Burst" is SEVEN cycyles of 3.57945 HZ in every line of video
to phase lock the crystal in the display, not every frame, or even every
field. The 'Tint' control trims the phase of the internal Chroma
oscillator, and it only takes a slight change in phase to make faces go
from green to purple. If it was off more than a cycle the chroma
unlocks, and you see slowly swirling color crap on the screen. Go more
than the 10 cycle limit and it triggers the 'color killer' circuit which
turns it to a mono display.

Ever heard of oven controlled crystal oscillators? The master sync
generator uses one, then 'Blackburst' is sent to every camera, VTR or
effects unit to provide the reference signal. They are powered
24/7/265.25 Even the series 300 Grass Valley group I used back in the
early '70s had dual generators with genlock and active failover.

All this was understood 60 years ago when NTSC approved the use of
Color TV in the US. They designed the system so the expensive hardware
was only needed at the station, and even that left the TVs at $1000+ in
the first couple years. Specially designed tubes allowed simpler
designs and fewer stages which helped, and that drop spurred more sales.
A nice color TV was still around $600 in the mid to late '60s.

 

[Dave Platt]

In article <bvidnesX7cyl4IDMnZ2dnUVZ_rudnZ2d@earthlink.com>,
Michael A. Terrell <mike.terrell@earthlink.net> wrote:

Ever heard of oven controlled crystal oscillators? The master sync
generator uses one, then 'Blackburst' is sent to every camera, VTR or
effects unit to provide the reference signal. They are powered
24/7/265.25

Summer rerun season being the exception? (>>grin<&lt

--
Dave Platt <dplatt@radagast.org> AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

 

[Michael A. Terrell]

Dave Platt wrote:

In article <bvidnesX7cyl4IDMnZ2dnUVZ_rudnZ2d@earthlink.com>,
Michael A. Terrell <mike.terrell@earthlink.net> wrote:

Ever heard of oven controlled crystal oscillators? The master sync
generator uses one, then 'Blackburst' is sent to every camera, VTR or
effects unit to provide the reference signal. They are powered
24/7/265.25

Summer rerun season being the exception? (>>grin<&lt

With 'syndication' it's always rerun season. A local station
shows the same episode of "Big Bang Theory" three times in one week.

 

On Thu, 14 Feb 2013 19:11:35 -0500, "Michael A. Terrell"
<mike.terrell@earthlink.net> wrote:

rickman wrote:

On 2/12/2013 12:25 PM, Michael A. Terrell wrote:

radams2000@gmail.com wrote:

Long ago I learned a painful lesson about video frequency accuracy. I assumed that getting a crystal within 100ppm of tbe ideal frequency should be just fine. But when we tried to send that signal through a satellite, it wouldn't go because the frequency accuracy was too poor.

I think the issue is that VCXO's are often used for clock recovery and they can't be pulled very far. Or maybe they just rely on absolute accuracy, and when things drift between the source and the sync, you just drop or repeat a frame. Not sure.


NTSC color calls for +/- 10 Hz for the 3,579,545 burst frequency.

Hmmm... ą3 ppm. Did they have much luck finding crystals that would
provide this? I expect there is more than just a crystal in this
circuit. I seem to recall that the color burst *is* a reference that
lets the local TV oscillator sync up on every frame. Maybe the camera
has a ą3 ppm oscillator in it, but I seriously doubt TVs do.


Apparently you've never looked at the schematic of a color TV or
monitor or the RS-170A standard.

"Color Burst" is SEVEN cycyles of 3.57945 HZ in every line of video
to phase lock the crystal in the display, not every frame, or even every
field. The 'Tint' control trims the phase of the internal Chroma
oscillator, and it only takes a slight change in phase to make faces go
from green to purple. If it was off more than a cycle the chroma
unlocks, and you see slowly swirling color crap on the screen. Go more
than the 10 cycle limit and it triggers the 'color killer' circuit which
turns it to a mono display.

Ever heard of oven controlled crystal oscillators? The master sync
generator uses one, then 'Blackburst' is sent to every camera, VTR or
effects unit to provide the reference signal. They are powered
24/7/265.25 Even the series 300 Grass Valley group I used back in the
early '70s had dual generators with genlock and active failover.

All this was understood 60 years ago when NTSC approved the use of
Color TV in the US. They designed the system so the expensive hardware
was only needed at the station, and even that left the TVs at $1000+ in
the first couple years. Specially designed tubes allowed simpler
designs and fewer stages which helped, and that drop spurred more sales.
A nice color TV was still around $600 in the mid to late '60s.

A nice color TV is still around $600. ;-) ...and probably always will
be.

 

On Thu, 14 Feb 2013 22:12:41 -0500, "Michael A. Terrell"
<mike.terrell@earthlink.net> wrote:

Dave Platt wrote:

In article <bvidnesX7cyl4IDMnZ2dnUVZ_rudnZ2d@earthlink.com>,
Michael A. Terrell <mike.terrell@earthlink.net> wrote:

Ever heard of oven controlled crystal oscillators? The master sync
generator uses one, then 'Blackburst' is sent to every camera, VTR or
effects unit to provide the reference signal. They are powered
24/7/265.25

Summer rerun season being the exception? (>>grin<&lt


With 'syndication' it's always rerun season. A local station
shows the same episode of "Big Bang Theory" three times in one week.

Looking at it another way, the term "season" really doesn't have much
meaning anymore.

 

[Phil Hobbs]

On 02/15/2013 05:11 PM, krw@attt.bizz wrote:

On Thu, 14 Feb 2013 19:11:35 -0500, "Michael A. Terrell"
mike.terrell@earthlink.net> wrote:


rickman wrote:

On 2/12/2013 12:25 PM, Michael A. Terrell wrote:

radams2000@gmail.com wrote:

Long ago I learned a painful lesson about video frequency accuracy. I assumed that getting a crystal within 100ppm of tbe ideal frequency should be just fine. But when we tried to send that signal through a satellite, it wouldn't go because the frequency accuracy was too poor.

I think the issue is that VCXO's are often used for clock recovery and they can't be pulled very far. Or maybe they just rely on absolute accuracy, and when things drift between the source and the sync, you just drop or repeat a frame. Not sure.


NTSC color calls for +/- 10 Hz for the 3,579,545 burst frequency.

Hmmm... ą3 ppm. Did they have much luck finding crystals that would
provide this? I expect there is more than just a crystal in this
circuit. I seem to recall that the color burst *is* a reference that
lets the local TV oscillator sync up on every frame. Maybe the camera
has a ą3 ppm oscillator in it, but I seriously doubt TVs do.


Apparently you've never looked at the schematic of a color TV or
monitor or the RS-170A standard.

"Color Burst" is SEVEN cycyles of 3.57945 HZ in every line of video
to phase lock the crystal in the display, not every frame, or even every
field. The 'Tint' control trims the phase of the internal Chroma
oscillator, and it only takes a slight change in phase to make faces go
from green to purple. If it was off more than a cycle the chroma
unlocks, and you see slowly swirling color crap on the screen. Go more
than the 10 cycle limit and it triggers the 'color killer' circuit which
turns it to a mono display.

Ever heard of oven controlled crystal oscillators? The master sync
generator uses one, then 'Blackburst' is sent to every camera, VTR or
effects unit to provide the reference signal. They are powered
24/7/265.25 Even the series 300 Grass Valley group I used back in the
early '70s had dual generators with genlock and active failover.

All this was understood 60 years ago when NTSC approved the use of
Color TV in the US. They designed the system so the expensive hardware
was only needed at the station, and even that left the TVs at $1000+ in
the first couple years. Specially designed tubes allowed simpler
designs and fewer stages which helped, and that drop spurred more sales.
A nice color TV was still around $600 in the mid to late '60s.

A nice color TV is still around $600. ;-) ...and probably always will
be.

Pournelle's Law still applies too: "The computer you want is always
$5000." Still true for me!

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

On Fri, 15 Feb 2013 17:14:12 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 02/15/2013 05:11 PM, krw@attt.bizz wrote:
On Thu, 14 Feb 2013 19:11:35 -0500, "Michael A. Terrell"
mike.terrell@earthlink.net> wrote:


rickman wrote:

On 2/12/2013 12:25 PM, Michael A. Terrell wrote:

radams2000@gmail.com wrote:

Long ago I learned a painful lesson about video frequency accuracy. I assumed that getting a crystal within 100ppm of tbe ideal frequency should be just fine. But when we tried to send that signal through a satellite, it wouldn't go because the frequency accuracy was too poor.

I think the issue is that VCXO's are often used for clock recovery and they can't be pulled very far. Or maybe they just rely on absolute accuracy, and when things drift between the source and the sync, you just drop or repeat a frame. Not sure.


NTSC color calls for +/- 10 Hz for the 3,579,545 burst frequency.

Hmmm... ą3 ppm. Did they have much luck finding crystals that would
provide this? I expect there is more than just a crystal in this
circuit. I seem to recall that the color burst *is* a reference that
lets the local TV oscillator sync up on every frame. Maybe the camera
has a ą3 ppm oscillator in it, but I seriously doubt TVs do.


Apparently you've never looked at the schematic of a color TV or
monitor or the RS-170A standard.

"Color Burst" is SEVEN cycyles of 3.57945 HZ in every line of video
to phase lock the crystal in the display, not every frame, or even every
field. The 'Tint' control trims the phase of the internal Chroma
oscillator, and it only takes a slight change in phase to make faces go
from green to purple. If it was off more than a cycle the chroma
unlocks, and you see slowly swirling color crap on the screen. Go more
than the 10 cycle limit and it triggers the 'color killer' circuit which
turns it to a mono display.

Ever heard of oven controlled crystal oscillators? The master sync
generator uses one, then 'Blackburst' is sent to every camera, VTR or
effects unit to provide the reference signal. They are powered
24/7/265.25 Even the series 300 Grass Valley group I used back in the
early '70s had dual generators with genlock and active failover.

All this was understood 60 years ago when NTSC approved the use of
Color TV in the US. They designed the system so the expensive hardware
was only needed at the station, and even that left the TVs at $1000+ in
the first couple years. Specially designed tubes allowed simpler
designs and fewer stages which helped, and that drop spurred more sales.
A nice color TV was still around $600 in the mid to late '60s.

A nice color TV is still around $600. ;-) ...and probably always will
be.


Pournelle's Law still applies too: "The computer you want is always
$5000." Still true for me!

I guess we have different wants. ;-) I thought I went nuts at $1300
(laptop/tablet convertible with multi-touch screen and dock). Don't
know how to add monitors into the cost though. They've been here
through three systems (and were only ~600 for the two).

 

[Michael A. Terrell]

krw@attt.bizz wrote:

On Thu, 14 Feb 2013 22:12:41 -0500, "Michael A. Terrell"
mike.terrell@earthlink.net> wrote:


Dave Platt wrote:

In article <bvidnesX7cyl4IDMnZ2dnUVZ_rudnZ2d@earthlink.com>,
Michael A. Terrell <mike.terrell@earthlink.net> wrote:

Ever heard of oven controlled crystal oscillators? The master sync
generator uses one, then 'Blackburst' is sent to every camera, VTR or
effects unit to provide the reference signal. They are powered
24/7/265.25

Summer rerun season being the exception? (>>grin<&lt


With 'syndication' it's always rerun season. A local station
shows the same episode of "Big Bang Theory" three times in one week.

Looking at it another way, the term "season" really doesn't have much
meaning anymore.

Considering that the new shows are presented raw?