Change input voltage polarity based on control signal

[Siddharth]

Hi all
I have 2 DC voltage inputs, Va and Vb. Va varies from 0V to 10V, while
Vb has output of + or - 1V.
I want to change the sign of Va depending on whether Vb is positive or
negative. All this has to be done in analog, not dsp.
Current through the circuit is 1A max. I've thought of using a Gilbert
2Q multiplier, but I was wondering if there was another way.
Thanks
Sid

 

[Tony Williams]

Hi all I have 2 DC voltage inputs, Va and Vb. Va varies from 0V
to 10V, while Vb has output of + or - 1V. I want to change the
sign of Va depending on whether Vb is positive or negative. All
this has to be done in analog, not dsp. Current through the
circuit is 1A max. I've thought of using a Gilbert 2Q multiplier,
but I was wondering if there was another way. Thanks Sid

There is a simple comparator-opamp circuit that
might do what you want.

ascii-art below, needs fixed-width font.

R R
Va-------------+----/\/\------+---/\/\--+
| | _ |
| +--|- \ |
| R2 |Op >--+-->Vout
+----/\/\------+--|+_/
_ |
Vb-----------|- \ +
|Co >---+------->/Switch
0v--/\/\---+-|+_/ | +
| | |
+--/\/\---+ --+--0v

The comparator senses the zero-crossing of Va, and
turns the Switch ON and OFF.

When Switch= ON, Vout = -Va, when OFF, Vout= +Va.

Operation is slightly improved if R2 and Switch are
replaced by a SPCO switch, switching the +ve input
between Va and 0v.

 

[Ken Taylor]

Hi all
I have 2 DC voltage inputs, Va and Vb. Va varies from 0V to 10V, while
Vb has output of + or - 1V.
I want to change the sign of Va depending on whether Vb is positive or
negative. All this has to be done in analog, not dsp.
Current through the circuit is 1A max. I've thought of using a Gilbert
2Q multiplier, but I was wondering if there was another way.
Thanks
Sid

Double pole change-over relay....?

Ken

 

[Jim Thompson]

Hi all
I have 2 DC voltage inputs, Va and Vb. Va varies from 0V to 10V, while
Vb has output of + or - 1V.
I want to change the sign of Va depending on whether Vb is positive or
negative. All this has to be done in analog, not dsp.
Current through the circuit is 1A max. I've thought of using a Gilbert
2Q multiplier, but I was wondering if there was another way.
Thanks
Sid

1A ??

There was a thread a while back (IIRC on A.B.S.E) that used power fets
to provide the right polarity to a load, irrespective of input
polarity.

I can't seem to track it down now, but perhaps some other poster can
recall it.

It would require a few additions to do what you want.

...Jim Thompson

 

[[email protected]]

How about an H-bridge chip, as long as you could get your +-1 volt
signal to look like a 1/0 logic input to the "direction" input.

 

[Joerg]

Hello,

How about an H-bridge chip, as long as you could get your +-1 volt
signal to look like a 1/0 logic input to the "direction" input.

That won't flip the polarity of Va unless Va was coming in decoupled
from ground.

Regards, Joerg

 

[Matrix]

Thanks everyone for your help. Tony, I dont think your circuit would
work. Notice when the switch is connected to Va, both inputs to the
opamp would be the same. For the input impedance to the opamp to be the
same, R2 = R. Thus, the voltage inputs to the terminals would be the
same.
Thanks though for your help, and please tell me if I'm wrong in my
thinking.
I think I'm going to out Va and -Va as inputs to a FET switch, with the
gate being the comparator output.
sid
P.S: I'm wondering how to conceal my email address to prevent spam. I
can't see a place to do that.

 

[Ban]

Thanks everyone for your help. Tony, I dont think your circuit would
work. Notice when the switch is connected to Va, both inputs to the
opamp would be the same. For the input impedance to the opamp to be
the same, R2 = R. Thus, the voltage inputs to the terminals would be
the same.

Nope. Tonys circuit will work, but not with 1A output current.
When the switch is open, Va is on the +in, -in and also the output. Va=Vout
when the switch is closed, +in is on ground, so the output must invert to
keep also -in on virtual ground. I think Matrix, you should read up on
opamps again.

ciao Ban

 

[Jim Thompson]

Thanks everyone for your help. [snip]
P.S: I'm wondering how to conceal my email address to prevent spam. I
can't see a place to do that.

Depends on newsreader. In Agent it's under "User and System Profile,
User"... just type in a munged E-mail address there.

...Jim Thompson

 

[Rich Grise]

P.S: I'm wondering how to conceal my email address to prevent spam. I
can't see a place to do that.

The "professional" answer to that is, of course, use a real newsreader
and ask your ISP for the name of their real newsserver, and avoid google
groups like the plague. (It used to be OK, until the google minions
found out that there's much more revenue to be had by underestimating
the intelligence of the public.)

What I did was make up a new bogus email at yahoo:
[email protected] . When I want to expose my real email, I say
"elide ard". I reckon that if someone wants to email me, they'll
bother to figure out what that means. Of course, as soon as I
used it for my googlegroups account, I went back to yahoo and
redirected _everything_ to the bit bucket. Damn! Sometimes I'm
so fuckin' clever I even scare myself! ;-P

Cheers!
Rich

 

[Rich Grise]

Thanks everyone for your help. [snip]
P.S: I'm wondering how to conceal my email address to prevent spam. I
can't see a place to do that.

Depends on newsreader. In Agent it's under "User and System Profile,
User"... just type in a munged E-mail address there.

JT, he's posting from googlegroups. They want a "real" email, so that
they can send you a "yes, I really signed up, and this is my real
email" autoresponse, and then they authorize you to post. At least,
that's the way it worked a few years ago. IOW, he's not even using
a newsreader - he's using a browser and google's atrocious chatroomoid
interface.

Sigh.

 

[JeffM]

I'm wondering how to conceal my email address to prevent spam.

sid (Matrix)

1) Never publish your primary email address on the Internet
unless it's munged
http://66.102.7.104/search?q=cache:R0u-koBQf6IJ:en.wikipedia.org/wiki/Mung+modify-an-e-mail-address
or unless (if you're not running your own email server)
your ISP's mail server has excellent anti-spam filtering.
Once your primary email address has gotten out into the wild,
set up a new email account and abandon the old one.

2) Before you give your primary email address to someone,
have that someone explain the difference between CC and BCC.
http://www.google.com/search?&q=BCC+blind-carbon-copy+CC+carbon-copy+russharvey

3) Never give your primary email address
to someone whom you would not trust with your wallet.

Set up throw-away email accounts for contact with everyone else.
http://www.spamgourmet.com/
http://www.emailaddresses.com/email_web.htm

Since you're already using hotmail, all you had to do
was create another account and abandon it after 1 use.
..
..
Rich already mentioned using a real newsreader:
http://www.google.com/search?&q=gravity+forte+xnews+40tude+mozilla

 

[legg]

There is a simple comparator-opamp circuit that
might do what you want.

ascii-art below, needs fixed-width font.

R R
Va-------------+----/\/\------+---/\/\--+
| | _ |
| +--|- \ |
| R2 |Op >--+-->Vout
+----/\/\------+--|+_/
_ |
Vb-----------|- \ +
|Co >---+------->/Switch
0v--/\/\---+-|+_/ | +
| | |
+--/\/\---+ --+--0v

The comparator senses the zero-crossing of Va, and
turns the Switch ON and OFF.

When Switch= ON, Vout = -Va, when OFF, Vout= +Va.

Operation is slightly improved if R2 and Switch are
replaced by a SPCO switch, switching the +ve input
between Va and 0v.

What do you suppose the OP expects to happen when Vb=0?

The intermediate values of resistance between open and short for the
switch do not produce gradual intermediate values.of gain.

RL

 

[Tony Williams]

[/QUOTE]

What do you suppose the OP expects to happen when Vb=0?

The intermediate values of resistance between open and short
for the switch do not produce gradual intermediate values.of
gain.

There's the OP's original Q above.

I read it that he wants a switch of polarity at Vb=0. Note
the allowance for a small amount of hysteresis (10mV?) around
the comparator.

 

[Fred Bartoli]

What do you suppose the OP expects to happen when Vb=0?

The intermediate values of resistance between open and short
for the switch do not produce gradual intermediate values.of
gain.

There's the OP's original Q above.

I read it that he wants a switch of polarity at Vb=0. Note
the allowance for a small amount of hysteresis (10mV?) around
the comparator.
[/QUOTE]

Tony,

the way I read it is that only 2 discrete values are possible for vb: + or -
1V.

In that case you can get rid of the comparator and just use a simple BJT.

 

[Tony Williams]

Tony, the way I read it is that only 2 discrete values are
possible for vb: + or - 1V.

In that case you can get rid of the comparator and just use a
simple BJT.

Ah yes. I can see that, now you have pointed it out.