I breadboard separate two transistor circuits to compares the
performance of switching in and out a 75 ohm resistor terminator to a
video source. One circuit is based on a bi-polar 2N3904 and the other
is a FET transistor(it's a DMOS N-channel MOSFET). Using a video
generator that is set to sweep 1 to 30MHz and an oscilloscope to
monitor the response, I observed that both type of terminators have
issues. On the bipolar method, the frequency response is pretty flat
to 30MHz when the 2n3904 is truned off, but when it is turned on, the
frequency response has gain peaks and loss valleys. This cause the
video sync to measure 40IRE (that's good), but the color burst
measures 48IRE (too high). I need the frequency response to be flat to
10 MHz. The FET mehtod has the reverse effect. When the FET is on,
the frequency response is pretty flat to 30MHz. When the FET is OFF,
the frequency response has a number of gain peaks and loss valleys.
Also, on both method, the video signal wobbles slowly slightly up and
down (on or foo on both methods). I can't seem to get rid of the
wobble. Must be due to the stray capacitance of both FET and bipolar.
Any help will be appreciated.


Circuit of both termination method:


video source------------------*-------------------Oscilloscope
|
C
0V or 5VDC--10k--------B 2N3904
E
|
|
75 ohm
|
|
|
-5VDC


video source------------------*-------------------Oscilloscope
|
D
0V or 5VDC--10k--------G N-channel DMOS FET RON(MAX)=2ohm
S for VGS= 5V
|
|
75 ohm
|
|
|
-5VDC

On Wed, 01 Feb 2006 21:55:47 -0500, LRCR wrote:

Also, on both method, the video signal wobbles slowly slightly up and
down (on or foo on both methods). I can't seem to get rid of the
wobble. Must be due to the stray capacitance of both FET and bipolar.
Any help will be appreciated.

Check the power supply ripple and look for ground loops or shielding
for the wobble. Can you see the wobble on the scope? If so, will it
sync to the power line?

I haven't tried switching in a resistor but from what little
experience I do have with video - it is very sensitive to layout and
switching speeds of components. I've had excellent results using CMOS
analog switches to change levels and switch video on and off -

If you're bread boarding with one of those plastic plug boards - you
may not get it to work - too much capacitance between tie points. A
ground plane can help some . . .

Check out the way a CD4066 is connected internally. Four cmos
transistors in the signal path both p and n channel and one more to
invert the control signal..
--

----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Sounds like an excellent idea. I will try it out. Thanks for the
CD4066 analogy. I will look into it. On the verticle wobble motion
of the video signal, I will use a regular scope and sync it to the
line to see the result.

Thanks
Caesar
On Thu, 02 Feb 2006 10:06:40 -0500, default wrote:

On Wed, 01 Feb 2006 21:55:47 -0500, LRCR wrote:

Also, on both method, the video signal wobbles slowly slightly up and
down (on or foo on both methods). I can't seem to get rid of the
wobble. Must be due to the stray capacitance of both FET and bipolar.
Any help will be appreciated.

Check the power supply ripple and look for ground loops or shielding
for the wobble. Can you see the wobble on the scope? If so, will it
sync to the power line?

I haven't tried switching in a resistor but from what little
experience I do have with video - it is very sensitive to layout and
switching speeds of components. I've had excellent results using CMOS
analog switches to change levels and switch video on and off -

If you're bread boarding with one of those plastic plug boards - you
may not get it to work - too much capacitance between tie points. A
ground plane can help some . . .

Check out the way a CD4066 is connected internally. Four cmos
transistors in the signal path both p and n channel and one more to
invert the control signal..

"LRCR" wrote in message
...
I breadboard separate two transistor circuits to compares the
performance of switching in and out a 75 ohm resistor terminator to a
video source. One circuit is based on a bi-polar 2N3904 and the other
is a FET transistor(it's a DMOS N-channel MOSFET). Using a video
generator that is set to sweep 1 to 30MHz and an oscilloscope to
monitor the response, I observed that both type of terminators have
issues. On the bipolar method, the frequency response is pretty flat
to 30MHz when the 2n3904 is truned off, but when it is turned on, the
frequency response has gain peaks and loss valleys. This cause the
video sync to measure 40IRE (that's good), but the color burst
measures 48IRE (too high). I need the frequency response to be flat to
10 MHz. The FET mehtod has the reverse effect. When the FET is on,
the frequency response is pretty flat to 30MHz. When the FET is OFF,
the frequency response has a number of gain peaks and loss valleys.
Also, on both method, the video signal wobbles slowly slightly up and
down (on or foo on both methods). I can't seem to get rid of the
wobble. Must be due to the stray capacitance of both FET and bipolar.
Any help will be appreciated.


Circuit of both termination method:


video source------------------*-------------------Oscilloscope
|
C
0V or 5VDC--10k--------B 2N3904
E
|
|
75 ohm
|
|
|
-5VDC


video source------------------*-------------------Oscilloscope
|
D
0V or 5VDC--10k--------G N-channel DMOS FET RON(MAX)=2ohm
S for VGS= 5V
|
|
75 ohm
|
|
|
-5VDC

I would try moving the transistor to the other side of the terminator so you
are switching -5VDC to the terminator rather than passing the video signal
thru the transistor. BTW I don't see how 0v on the base of an npn when the
emitter is at -5 (thru 75 ohms) will turn it off.