EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Unixon on February 05, 2014, 08:43:46 pm
-
Hi! Need a little advice on the topic.
I have a circuit that outputs a signal withing range 0V .. Vcc around the quiescent value Vref = Vcc/2 and I need to convert this to absolute value in the same range (0 .. Vcc) as the original signal. Most of schematics for absolute value amplifiers I found uses two opamps and suggests that output range is only a half Vcc/2 .. Vcc of a full power supply range. I tried to figure out how to redesign these circuits to output at full range using just the same two opamps, but couldn't come to a solution in short time. The PCB area in this project is very limited and I can't go on with adding another IC package.
Can anyone suggest a modification of either of following circuits or a an alternative schematic?
http://www.cirvirlab.com/index.php/tutorials/77-op-amp-absolute-value-amplifier-with-one-diode.html (http://www.cirvirlab.com/index.php/tutorials/77-op-amp-absolute-value-amplifier-with-one-diode.html)
http://www.linear.com/solutions/1608 (http://www.linear.com/solutions/1608)
I know there are lot other arrangements for such circuit, these two are provided just for the reference.
-
Up
-
Your not clear on what the transformation needs to be, it sounds like your signal is +-Vcc/2 offset by Vcc/2, and you want to represent that as +-Vcc?
sas such am left wondering why a differential amplifier with a varied gain would not work?
-
Isn't what you are suggesting just rectifcation?
If your output is only 0 to Vcc/2 why not just add a x2 gain stage
-
Yes, it is a rectification, but I want the signal to be rectified and amplified at the same time.
Vin varies from 0V to Vcc and when there's no meaningful signal Vin settles at Vref=Vcc/2.
Vout varies from Vcc/2 to Vcc in the original scheme and I want Vout to vary from 0V to Vcc.
The problem is that I can't just drop in another amplification stage which would be perfectly easy way to go.
I need first two opamps to produce this x2 gain along with rectification.
So I'm looking for a configuration of feedback circuit that would allow to do this rectification+amplification with only two opamps.
-
This is one of those trick circuit app notes that I always remember but haven't had an actual use for yet. Check it out.
http://www.analog.com/library/analogdialogue/archives/44-04/absolute.html (http://www.analog.com/library/analogdialogue/archives/44-04/absolute.html)
-
Yes, I've seen this before. A little different arrangement than what I had provided for the reference, but just the same result.
I found required configurations for positive and negative signals (relative to Vref) separately, but not for complete signal.
-
I know there a lots of sneaky little circuit tricks around but I'm struggling to see what you want achieved without an extra opamp and a negative supply.
If you do find a solution please post it here
-
Adding an extra opamp is easy, but I want to squeeze this easy circuit into only two opamps.
The reason is available PCB area. I can afford one SO8 or a couple of SOT23-5 packages,
but adding an extra opamp will force me to switch to SO14/SO16 or add an extra SOT23 package.
I basically have no space for this including feedback resistors that will eat free space even faster.
As for the signal itself,
I want to achieve twice as better scale in further processing by rectifying it with x2 gain.
All circuits are unipolar, working voltage is 5V (3.3V should be also OK).
Maybe I should look for even tinier packages like SC-70 and switch to 0402 for resistors (from 0603)...
Fortunately, I will be able to afford adding an extra opamp... not sure if I should proceed this way.