4066 bandwidth up to almost 100MHz?

[Deefoo]

I want to make a simple numerically-controlled attenuator for a 6Vtt signal
with a bandwidth of 20MHz. Looking at analog switches I was wondering if I
could simply use a classic 4066 for this task. Philips state an on-state
frequency response of 90MHz for their HEF4066B but they don't show any
graphs. Would this work? They sure are cheap compared to other switches.

Thanks,
--DF

 

[Winfield Hill]

Deefoo wrote...

I want to make a simple numerically-controlled attenuator for a 6Vtt signal
with a bandwidth of 20MHz. Looking at analog switches I was wondering if I
could simply use a classic 4066 for this task. Philips state an on-state
frequency response of 90MHz for their HEF4066B but they don't show any
graphs. Would this work? They sure are cheap compared to other switches.

I prefer 74hc4053 which allow for making RF attenuators with T-style
switching. The concern is not only with bandwidth, but feedthough.

 

[Deefoo]

Deefoo wrote...
I prefer 74hc4053 which allow for making RF attenuators with T-style
switching. The concern is not only with bandwidth, but feedthough.

I am not sure I am getting your drift. Do you mean bypassing or not the
T-network? Then why not a PI-network? Or do you have something else in mind?

Thanks,
--DF

 

[Jim Thompson]

Deefoo wrote...
I prefer 74hc4053 which allow for making RF attenuators with T-style
switching. The concern is not only with bandwidth, but feedthough.

Win, What off-the-shelf device would you use to multiplex two video
signals, both have the same sync, and I want to switch frame-by-frame?

...Jim Thompson

 

[Joerg]

Hello Jim,

Win, What off-the-shelf device would you use to multiplex two video
signals, both have the same sync, and I want to switch frame-by-frame?

I used to do that with the FET quad array SD5400 but that seems to have
fallen from grace by now.

Check out the THS73xx series, mainly the THS7313. Has a 2:1 input mux.
Guess you don't need three channels but for a little over a buck, why not?

I don't know if the mux can be switched fast enough via I2C to do it in
one V-sync period. The 7313 supports the 400kbps mode. Supposedly works
in I2C high speed (3.4Mbps) but AFAIK that's not guaranteed.

 

[Joerg]

Check out the THS73xx series, mainly the THS7313. ...

Forgot to say: These are from Texas Instruments.

 

[Jim Thompson]

Hello Jim,


I used to do that with the FET quad array SD5400 but that seems to have
fallen from grace by now.

Check out the THS73xx series, mainly the THS7313. Has a 2:1 input mux.
Guess you don't need three channels but for a little over a buck, why not?

I saw those in the latest Electronic Design.

I don't know if the mux can be switched fast enough via I2C to do it in
one V-sync period. The 7313 supports the 400kbps mode. Supposedly works
in I2C high speed (3.4Mbps) but AFAIK that's not guaranteed.

I may just do a "Joerg" and roll my own with discrete transistors.

Thinking back to the '60's I did an ultra-fast sample-and-hold with a
ECL-based tri-stateable unity-gain amplifier (my patent 3,643,110).

IIRC I could do sub-microsecond switching.

...Jim Thompson

 

[Joerg]

Hello Jim,

I may just do a "Joerg" and roll my own with discrete transistors.

Thinking back to the '60's I did an ultra-fast sample-and-hold with a
ECL-based tri-stateable unity-gain amplifier (my patent 3,643,110).

IIRC I could do sub-microsecond switching.

When it has to be blazingly fast you can do the diode quad and toroid.
The first time I did that was for my final project back at the
university. I needed three samplers to hit the sweet spots on a CCD
readout register. The sampling window had to be 15nsec and you couldn't
be off more than a couple nsec in timing, and absolutely no jitter. IIRC
I used 74AS logic as a drive, plus "analog slope enhancers". The toroids
were, of course, hand wound. Didn't want to wait for a Mini Circuits
order and wanted to stay within budget. The folks at that institute were
a bit baffled when I asked for the budget the project would have. Guess
nobody ever did that before. Got into trouble anyway because I spent
half that budget on phone calls and telex charges (remember those?) just
to find all the parts I needed. But, I stayed about a hundred under the
grand total.

 

[Winfield Hill]

Jim Thompson wrote...

Win, What off-the-shelf device would you use to multiplex two video
signals, both have the same sync, and I want to switch frame-by-frame?

Don't know of a way to do it with one IC. But you know about
the several nice miniDIP sync-detector chips. The ones I've
used are spot on for timing. You could drive a flipflop with
the sync to get a select signal. I'd buffer the video first
to lower the signal impedance, and use a 74hc4053 to make the
signal selection, with a G=2 follower amp after, plus a 75R
resistor for the video out. You may need to have capacitive
coupling and black-level restoration in the mix, but this can
be done with the unused 4053 sections, using a set of timing
pulses derived from the sync chip. I don't recall the exact
details just now, but my old design drawings are in a folder.

 

[Jim Thompson]

Jim Thompson wrote...

Don't know of a way to do it with one IC. But you know about
the several nice miniDIP sync-detector chips. The ones I've
used are spot on for timing. You could drive a flipflop with
the sync to get a select signal. I'd buffer the video first
to lower the signal impedance, and use a 74hc4053 to make the
signal selection, with a G=2 follower amp after, plus a 75R
resistor for the video out. You may need to have capacitive
coupling and black-level restoration in the mix, but this can
be done with the unused 4053 sections, using a set of timing
pulses derived from the sync chip. I don't recall the exact
details just now, but my old design drawings are in a folder.

I have a whole library of DC restoration schemes

I just thought there might be a fast multiplexor out there that I'm
not aware of... I haven't done video, even for my own G-jobs, for
close to 20 years.

...Jim Thompson

 

[Joerg]

Hello Jim,

I just thought there might be a fast multiplexor out there that I'm
not aware of... I haven't done video, even for my own G-jobs, for
close to 20 years.

This here paramator indexor has multiplexors:
http://para.maxim-ic.com/cache/en/results/5148.html

For a one-up project it may be ok. Else you'd have to find out whether
it's available in qties and for how long. The latter can make for a
white knuckle ride.

 

[Kevin White]

I want to make a simple numerically-controlled attenuator for a 6Vtt signal
with a bandwidth of 20MHz. Looking at analog switches I was wondering if I
could simply use a classic 4066 for this task. Philips state an on-state
frequency response of 90MHz for their HEF4066B but they don't show any
graphs. Would this work? They sure are cheap compared to other switches.

Thanks,
--DF

Pericom has some good multiplexers for video switching - e.g. P15V331Q.

They also look useful for other things than video as they have have
fast turn-on/turn-off times and good capacitance to ground parameters,
most other fast multiplexers have horrendous capacitance to ground -
presumably because they use large devices to get the on-resistance low.
I was recently looking for something along the lines of the HC4053 but
with fast switching times and this was the closest I could find - it
does only a limited input voltage range though as it is intended for
video.

It is available from Digikey although their search engine often cannot
find it!!!
(PI5V331Q-ND)

I find the classic 4016/4066 is inconvenient to use because it does not
have level shifters on the logic control lines although for video it
would be OK because of the limited voltage excursion of video (0-1V).

cheers

Kevin

 

[John Jardine.]

I have a whole library of DC restoration schemes

I just thought there might be a fast multiplexor out there that I'm
not aware of... I haven't done video, even for my own G-jobs, for
close to 20 years.

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice480)460-2350 | |
| E-mail Address at Website Fax480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |

I love to cook with wine. Sometimes I even put it in the food

The TSH94 video amp will do wired OR. 200nS switch speed.
john

 

[Winfield Hill]

Jim Thompson wrote...

I have a whole library of DC restoration schemes

I just thought there might be a fast multiplexor out there that
I'm not aware of... I haven't done video, even for my own G-jobs,
for close to 20 years.

There are a variety of switch-amplifier combinations, as others
have said, but I like the hc4053 with a roll-your-own amplifier.
It's hard to find a faster, lower-capacitance switch, certainly
not a cheaper, easier-to-get one! It fills the bill for video.

 

[Fred Bloggs]

Jim Thompson wrote...



There are a variety of switch-amplifier combinations, as others
have said, but I like the hc4053 with a roll-your-own amplifier.
It's hard to find a faster, lower-capacitance switch, certainly
not a cheaper, easier-to-get one! It fills the bill for video.

It must have lousy distortion specs or something if the specialized
video muxes exist.

 

[Jim Thompson]

message news:[email protected]... [snip]

I just thought there might be a fast multiplexor out there that I'm
not aware of... I haven't done video, even for my own G-jobs, for
close to 20 years.

...Jim Thompson

The TSH94 video amp will do wired OR. 200nS switch speed.
john

That looks good. Thanks, John!

...Jim Thompson

 

[Michael A. Terrell]

Jim Thompson wrote...

Don't know of a way to do it with one IC. But you know about
the several nice miniDIP sync-detector chips. The ones I've
used are spot on for timing. You could drive a flipflop with
the sync to get a select signal. I'd buffer the video first
to lower the signal impedance, and use a 74hc4053 to make the
signal selection, with a G=2 follower amp after, plus a 75R
resistor for the video out. You may need to have capacitive
coupling and black-level restoration in the mix, but this can
be done with the unused 4053 sections, using a set of timing
pulses derived from the sync chip. I don't recall the exact
details just now, but my old design drawings are in a folder.

http://www.national.com/pf/LM/LM1881.html
--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[martin griffith]

Win, What off-the-shelf device would you use to multiplex two video
signals, both have the same sync, and I want to switch frame-by-frame?

...Jim Thompson

Almost anything would do, as long as it is glitch free after the
syncs, a couple of anything from maxim's stable would be OK, as long
as the diff gain/phase specs are ok
Switching video in analogue is dying out
Personally I'm glad to see the end of PAL/NTSC based equipment, a pain
in the arse in post production, great for transmission though, well
thought out, by the time they got PAL working


martin

 

[John Jardine.]

[...]

Switching video in analogue is dying out
Personally I'm glad to see the end of PAL/NTSC based equipment, a pain
in the arse in post production, great for transmission though, well
thought out, by the time they got PAL working


martin

Yes. A lot of tasty digital kit is now available for the high quality studio
end.
I must mention, that at the pleb end, i.e. the burgeoning UK video security
industry. It still relies near totally on analogue composite PAL. Indeed,
it's difficult to buy a (good!) security cam that offers digital out.
Stick a PAL cam on a pole, send the 6MHz balanced/unbalanced off for
monitoring and the car park riff-raff are still clearly identifiable.
Digitise and compress it a little and you're left with a nice pretty
picture but near useless for the intended purpose.
(I'm biased. I earn money from designing analogue matrix routers .
john

 

[Hal Murray]

Check out the THS73xx series, mainly the THS7313. Has a 2:1 input mux.
Guess you don't need three channels but for a little over a buck, why not?

I don't know if the mux can be switched fast enough via I2C to do it in
one V-sync period. The 7313 supports the 400kbps mode. Supposedly works
in I2C high speed (3.4Mbps) but AFAIK that's not guaranteed.

Does the I2C command have to be fast, or just the last clock happen
at the right time?