inverting amplifier and voltage divider

Status
Not open for further replies.
M

menewbie

Junior Member level 1
Joined
Mar 12, 2006
Messages
17
Helped
2
Reputation
4
Reaction score
0
Trophy points
1,281
Activity points
1,443
Hi, is there a circuit that can invert the sign of only positive signal? For example, if the input is 3V then the output is -3V and if input is -2V then the output is -2V.

Also, how to connect the output of a voltage divider to an integrator??
I am using 10p capacitors to divide the voltage but when I connect it to the integrator, which consists of an op-amp, a resistor(1k) and a capacitor(10u), the output of the voltage divider becomes 0V. I know I an impedance transformer can solve this problem. However, I am not sure what are the right values for the input resistance and the output resistance.
 

For your first question, do you mind to have a diode voltage drop on the output? If not, you can achieve the goal by a diode bridge, which works like a "negative" rectifier.
 

yes,
they are called polarity reversal circuits.
u can go thru the following link. it has got the block diagram of polarity reversal circuit.
**broken link removed**
 

jlee said:
For your first question, do you mind to have a diode voltage drop on the output? If not, you can achieve the goal by a diode bridge, which works like a "negative" rectifier.
Click to expand...

Hmm, the accuracy is very important for the circuit I am trying to design. If I have a diode voltage drop on the output then the accuracy will suffer....is there a way to compensate for that diode voltage drop?? I need accuracy up to 10uV
 


For the first question, Maybe you need the combination of inverting amplifier, non-inverting amplifier, comparator and switches to realize this function.
For the second question, can give output your signal frequency?

sixth
 

Chethan said:
yes,
they are called polarity reversal circuits.
u can go thru the following link. it has got the block diagram of polarity reversal circuit.
h**p://sound.westhost.com/project90.htm
Click to expand...

Thank you for the reply but the accuracy is very important in the circuit that I am trying to design. The circuit shown on the webpage uses resistors and I think for a -5V signal, the circuit will give a 5V signal. I want the output to be a -5V signal when the input is 5V.
 

inverter

yes it's name is rectifier.
see this link:
**broken link removed**

regards,
Davood.
 

Re: inverter

sixth said:
For the first question, Maybe you need the combination of inverting amplifier, non-inverting amplifier, comparator and switches to realize this function.
For the second question, can give output your signal frequency?

sixth
Click to expand...

I see...yeah if i use inverting amplifier, non-inverting amplifier, comparator and switches then the area of the circuit will increase to a point that I cannot not fit the whole circuit into a chip.

I am operating the circuit under a low-frequency environment

Added after 5 minutes:

Davood Amerion said:
yes it's name is rectifier.
see this link:
h**p://www.play-hookey.com/analog/full-wave_rectifier.html

regards,
Davood.
Click to expand...

I see...thank you for the reply. I forgot to mention that accuracy is very important for my design. So using a rectifier would be harmful to the accuracy since the resistance varies 20% in the real world....
 

Re: inverter

menewbie said:
... So using a rectifier would be harmful to the accuracy since the resistance varies 20% in the real world....
Click to expand...
20% !!!????
 

I don't think you even checked the link Davood gave you, did you?
It was a good reference, and since I didn't think you looked, I copied a line of text from the site:

It is not just a diode bridge. It will always put out a negative voltage, and is probably very accurate. I'm sure it'll be just as accurate as your next stage. Just go have a look, I think it'll serve you.
Best wishes,
Robert
 


I dont think you understand what I meant in my reply. I checked the link. The rectifier uses resistors. In real world, a 5k resistor does not really have resistance of 5000 ohms... Instead, it might have a resistance in the range of 4000ohms to 6000ohms depedns on typs of the resistor . So if I use that rectifier, the accuracy of the whole circuit will suffer.
 

Yes, but not all resistors are 5%. Some are 10% (silver band?) some are 5% (gold) and you can get 1% which usually have 5 bands. So in a 5K resistor that amounts to plus or minus 500 ohms. What you could do is measure each of your precision resistors with a good ohm meter, and find two that are very nearly identical. This should give you what you need, no? Or you could do something like this with an instrument amp, which has very precise laser cut resistors, but usually costs around $20.
I couldn't stop thinking last night about what it is, exactly, that you are doing. That might help people help you.
Regards,
Robert
 


As I mentioned before, I am designing this circuit on a chip level and accuracy is very important. Even 1% can be extremely harmful to the circuit (ADC) that I am trying to design.
 

what is it topic for?
it seems menewbie (just) going to get points!
he consumes his point as soon as he get it!!!!
is'nt it?
 

menewbie,
The precision full wave rectifier shown at the website suggested by Dawood does exacly what you want. Preecision resistor networks are available with ratios matched to .01%.
Regards,
Kral
 

Kral said:
menewbie,
The precision full wave rectifier shown at the website suggested by Dawood does exacly what you want. Preecision resistor networks are available with ratios matched to .01%.
Regards,
Kral
Click to expand...

Yeah I know they are available. Like I said before, I need accuracy up to 10uV. Let's assume I have a 1.5V input then the output has to be -3V exactly. Let the values of resistors be 20k and 10k.
20k * 0.0001 = 2. so the actual resistance varies from 19998 to 20002 for the 20k resistor. for 10k resistor, the resistance is in the range of 9999 to 10001.
20002/ 9999 = 2.00040004 (The max ratio of the resistors)
19998/ 9999 = 1.9996 (The min ratio of the resistors)
Take the max ratio and mutliply it by the input voltage by 2.
2.0004 *1.5 = 3.0006
It is obvious that the accuracy is only up to 1mV and NOT 10uV.

Look, I know the circuit does exactly what I want but the accuracy is just simply not good enough.

Also, I already mentioned it many times before...I am designing this on chip level...Precision resistors are very hard to make and very expensive
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…