VEE on CMOS IC

[Jack B. Pollack]

I am trying to build a project using a CD4051B demultiplexor
(http://www.ee.washington.edu/stores/DataSheets/cd4000/cd4051.pdf). The
output of the chip is fairly inconsistent, some times it outputs other times
it does not. If a wave my had over it sometimes it works. If I ground VEE
sometimes it works. I think my problem is with the VEE pin voltage.

The way I have things wired is as follows:

VDD = +5V
VSS = GND
IHN = L
VEE (see below)

I have tried grounding VEE, grounding it through a 10K resistor, and
grounding it through a small cap. None of these thing made the circuit more
stable.

What do I want to use a VEE? Any other ideas on how to make the circuit more
stable.

Thanks

 

[John Larkin]

I am trying to build a project using a CD4051B demultiplexor
(http://www.ee.washington.edu/stores/DataSheets/cd4000/cd4051.pdf). The
output of the chip is fairly inconsistent, some times it outputs other times
it does not. If a wave my had over it sometimes it works. If I ground VEE
sometimes it works. I think my problem is with the VEE pin voltage.

The way I have things wired is as follows:

VDD = +5V
VSS = GND
IHN = L
VEE (see below)

I have tried grounding VEE, grounding it through a 10K resistor, and
grounding it through a small cap. None of these thing made the circuit more
stable.

What do I want to use a VEE? Any other ideas on how to make the circuit more
stable.

Thanks

What's your analog signal swing range?

John

 

[Jim Thompson]

I am trying to build a project using a CD4051B demultiplexor
(http://www.ee.washington.edu/stores/DataSheets/cd4000/cd4051.pdf). The
output of the chip is fairly inconsistent, some times it outputs other times
it does not. If a wave my had over it sometimes it works. If I ground VEE
sometimes it works. I think my problem is with the VEE pin voltage.

The way I have things wired is as follows:

VDD = +5V
VSS = GND
IHN = L
VEE (see below)

I have tried grounding VEE, grounding it through a 10K resistor, and
grounding it through a small cap. None of these thing made the circuit more
stable.

What do I want to use a VEE? Any other ideas on how to make the circuit more
stable.

Thanks

What is the range of your ANALOG signal?

Required (for 5V control logic):

VDD = +5V

VSS (called GND on many data sheets) = 0

VEE = More negative than most negative ANALOG input signal.

You might want to try the 74HC4051 series of multiplexors...
particularly if you choose the ON Semiconductor product it will have
lower Ron, and better linearity.

(When I redesigned that series of devices for ON Semiconductor to fit
a different manufacturing process I improved several performance
specs, Ron, Ron linearity versus analog signal voltage, switching
speed and charge injection

...Jim Thompson

 

[John Larkin]

(When I redesigned that series of devices for ON Semiconductor to fit
a different manufacturing process I improved several performance
specs, Ron, Ron linearity versus analog signal voltage, switching
speed and charge injection

...Jim Thompson

Thanks... I'll try them. HC40xx parts generally have huge charge
injections, enough to do wild things to the output of most opamps.

Can you recall how many pC?

John

 

[Kevin Aylward]

What is the range of your ANALOG signal?

Required (for 5V control logic):

VDD = +5V

VSS (called GND on many data sheets) = 0

VEE = More negative than most negative ANALOG input signal.

You might want to try the 74HC4051 series of multiplexors...
particularly if you choose the ON Semiconductor product it will have
lower Ron, and better linearity.

(When I redesigned that series of devices for ON Semiconductor to fit
a different manufacturing process I improved several performance
specs, Ron, Ron linearity versus analog signal voltage, switching
speed and charge injection

Ok...other then using the mosfets as designed by the processes guys, as
switches, what exactly did *you* do to improve performance?

Kevin Aylward
[email protected]
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

 

[Jim Thompson]

Jim Thompson wrote: [snip]

You might want to try the 74HC4051 series of multiplexors...
particularly if you choose the ON Semiconductor product it will have
lower Ron, and better linearity.

(When I redesigned that series of devices for ON Semiconductor to fit
a different manufacturing process I improved several performance
specs, Ron, Ron linearity versus analog signal voltage, switching
speed and charge injection

Ok...other then using the mosfets as designed by the processes guys, as
switches, what exactly did *you* do to improve performance?

Kevin Aylward

[snip]

None of your business, Mr. Nobody ;-)

I basically did a total "clean" design, ignoring the existing
schematics, and also did my own device-sizing/fingers, compensating
devices, etc.

In the ON Semi design review presentation they were stunned at how
much better linearity I attained.

...Jim Thompson

 

[John Woodgate]

I read in sci.electronics.design that Kevin Aylward

Ok...other then using the mosfets as designed by the processes guys, as
switches, what exactly did *you* do to improve performance?

Da doo Ron Ron Ron, da doo Ron Ron?

 

[Jim Thompson]

I read in sci.electronics.design that Kevin Aylward


Da doo Ron Ron Ron, da doo Ron Ron?

Kevin is becoming almost as intolerable as Burridge. Just about time
to PLONK him ;-)

...Jim Thompson

 

[Jack B. Pollack]

I want to use the CD4051B as an analog switch to "physically" connect 2 pins
of another IC together.

I have a IR remote control that I want to be able to fake key-presses on.
I've wired 2 CD4051B into the keyboard matrix with their common in/out
lines tied together. The cannel in/out pins are wired to the remote
control's IC pins.


I don't know what my analog signal swing range is. I also didn't understand
the rest of the threads of this post.

Thanks

 

[John Larkin]

I basically did a total "clean" design, ignoring the existing
schematics, and also did my own device-sizing/fingers, compensating
devices, etc.

In the ON Semi design review presentation they were stunned at how
much better linearity I attained.

...Jim Thompson

Yeah, stunning people is one of my favorite sports. You ask a few
basic questions, let them talk, sorta keep quiet and humble and
low-key for a while... then BANG! I think the stun-gun approach to
presenting technology is useful in disarming/squashing any in-house
competitors. If you come on too strong at first, they get their
defenses up. If you sneak in under the radar and drop all your bombs
at once, you win.

John

 

[Frank Bemelman]

I am trying to build a project using a CD4051B demultiplexor
(http://www.ee.washington.edu/stores/DataSheets/cd4000/cd4051.pdf). The
output of the chip is fairly inconsistent, some times it outputs other times
it does not. If a wave my had over it sometimes it works. If I ground VEE
sometimes it works. I think my problem is with the VEE pin voltage.

The way I have things wired is as follows:

VDD = +5V
VSS = GND
IHN = L
VEE (see below)

I have tried grounding VEE, grounding it through a 10K resistor, and
grounding it through a small cap. None of these thing made the circuit more
stable.

What do I want to use a VEE? Any other ideas on how to make the circuit more
stable.

Using this to fake keypresses on your remote, you have to connect VEE to
ground,
and also to the minus of the remote control battery.

Then select the row and column using the A/B/C inputs of each 4051. Check if
one 4051 is only connected to rows and the other only to columns.

To perform the actual fake keypress, you need a low pulse on INH, normally
it
has to be high. You can use a pullup resistor, 10K from INH to +5V and a
pushbutton from INH to ground. Pressing the button will connect the selected
row to the selected column.

 

[Jim Thompson]

Yeah, stunning people is one of my favorite sports. You ask a few
basic questions, let them talk, sorta keep quiet and humble and
low-key for a while... then BANG! I think the stun-gun approach to
presenting technology is useful in disarming/squashing any in-house
competitors. If you come on too strong at first, they get their
defenses up. If you sneak in under the radar and drop all your bombs
at once, you win.

John

I even pulled the "stun-gun" approach off successfully in a meeting
with Sanyo... typical Japanese... bunch of **really** old-farts... I
bowed a lot, sat quietly thru their presentation, then walked to the
board and KABAMM!

I was awarded the project.

...Jim Thompson

 

[John Larkin]

I even pulled the "stun-gun" approach off successfully in a meeting
with Sanyo... typical Japanese... bunch of **really** old-farts... I
bowed a lot, sat quietly thru their presentation, then walked to the
board and KABAMM!

I was awarded the project.

...Jim Thompson

Damn, but that's fun! I did a couple of PhD chemists just yesterday.
Now they're practically begging me to take their money, just because
none of them can do the numbers on opamps.

"Can you measure 200 nanovolts?"

"200 nv? That's a heap of signal!"

The Japanese seem to be impressed by American mavericks, but you do
have to be polite and respectful too.


John

 

[nospam]

I want to use the CD4051B as an analog switch to "physically" connect 2 pins
of another IC together.

I have a IR remote control that I want to be able to fake key-presses on.
I've wired 2 CD4051B into the keyboard matrix with their common in/out
lines tied together. The cannel in/out pins are wired to the remote
control's IC pins.

Now we get the rest of the story.

My guess is you have only made connections between the remote keypad and
the 4051 0,,7 inputs. That leaves the signals you are trying to switch
just floating around with respect to the 4051 supplies. Connect ground on
the remote to the 4051 ground. VEE on the 4051 should also be ground.

 

[Jack B. Pollack]

Your suggestion is pretty much what I did. The circuit is not stable,
sometimes I can get the remote to recognize the inputs from the 4051's and
other times I can't. It can work perfectly for 10 minutes and then not work
at all. resetting power doesn't make a difference. It will then start to
work again later.

 

[Jack B. Pollack]

All GNDs are tied together.

 

[Jonathan Kirwan]

(When I redesigned that series of devices for ON Semiconductor to fit
a different manufacturing process I improved several performance
specs, Ron, Ron linearity versus analog signal voltage, switching
speed and charge injection

I can't even find the charge injection spec on the datasheet.

My hobby interest in analog switches is more with very high Roff and having both
balanced as well as low charge injection for use with integrators for currents
<< 1uA (Ron is less important there.) What's good on those scores,
semiconductor-wise?

Jon

 

[John Larkin]

I can't even find the charge injection spec on the datasheet.

My hobby interest in analog switches is more with very high Roff and having both
balanced as well as low charge injection for use with integrators for currents
<< 1uA (Ron is less important there.) What's good on those scores,
semiconductor-wise?

Jon

MAX338 has very low injection and leakage, but you pay the price of
high Ron.

On many parts, charge injection crosses zero at some signal-voltage
level, which implies that you can skew the power supply voltages to
put your operating point (say, zero volts to reset an integrator) at
the zero-injection point. Well under 1 pC should be possible doing
this.

The ultimate integrator reset is supposed to be a photosensitive fet,
driven by an LED. It would be interesting to try some of the photofet
solid-state relays, too.

John

 

[Pooh Bear]

I want to use the CD4051B as an analog switch to "physically" connect 2 pins
of another IC together.

I have a IR remote control that I want to be able to fake key-presses on.
I've wired 2 CD4051B into the keyboard matrix with their common in/out
lines tied together. The cannel in/out pins are wired to the remote
control's IC pins.

Didn't you post about this a few weeks back ?

I don't know what my analog signal swing range is.

It's the voltage that appears on the pins of the remote control IC. Assuming
that the remote control runs off a pair of AA or AAA cells then 0-3V is the
answer.

I also didn't understand the rest of the threads of this post.

That's 'cos you're not a circuit designer. It's not meant to be easy. That's why
we study to be designers.

Why did you use the 4051 ?

The CD 4016 or 4066 make for simpler implementations of 'crosspoints'.

Did you connect your 4051 ground ( i.e. Vss ) to the negative of the remote
control IC ? That is *highly* important or there's no common ground.


Graham

 

[Winfield Hill]

Jack B. Pollack wrote...

Your suggestion is pretty much what I did. The circuit is not stable,
sometimes I can get the remote to recognize the inputs from the 4051's and
other times I can't. It can work perfectly for 10 minutes and then not work
at all. resetting power doesn't make a difference. It will then start to
work again later.

Sounds like a floating logic input pin.