Input stage for ADC with operational amplifier

I am trying to make a voltmeter with a 10-bit ADC. The ADC can measure only positive voltages so I have made an input stage with Operational Amplifiers.

The idea of the circuit is to transform any negative voltage to positive. V1 is the voltage output and V Detect is 3.3V or 0V if the input signal is negative or positive, respectively.

The input frequency will not be more than 200 KHz. Will this circuit work ok whit the TL084? Is there any special parameter I should take into account?

Also I need V1 not to exceed 3.3V, will a zener diode work or it will cause any problems with regards to capacitance? I know ADC are very sensitive to any external capacitance because of the internal sample-hold capacitor they have.

Some observations:

1. The two middle opamps don't have power. Opamps need power to function according to spec.

2. Actually after Steven's comment I see all 4 opamps are in a single package, which apparently is powered. The problem is therefore one of bad schematic drawing. When you have multiple parts in a single package with common power like this, usually you show the power pins as a separate schematic part, like U13E for example. You can show the power and ground pins on one of the four opamps, but that can add clutter and makes that amp look more special than the others. However, never do what you did, which is to show the power and ground pins on each schematic part.

Even worse, you then showed two of the parts with power and ground connected, and two without. What if the two pairs were connected to different power and ground voltages? That would be really misleading since it's not physically possible. This kind of schematic convention makes no sense and invites exactly the kind of confusion shown in point 1 above. I left that there so you can see what the first reaction to this misleading schematic was.

3. There is no bypass cap anywhere accross the opamp power pins.

4. You don't say what voltage Vcc is. The TL084 needs a few volts headroom from each power rail, so Vcc should be at least the largest signal voltage plus the headroom.

5. 220 Ω seems very small for R33. Can the TL084 really drive that much current (I haven't checked)? Even if it can, why is that necessary?

6. With D13 in series with R33, Vdetect will be undriven (high impedance) when the opamp goes low. Is this really what you want?

7. D17 shouldn't be right on the output of U13C. That will violate the max output current source capability of the opamp when it is trying to drive high.

• About 1.: it's worse. They're 4 opamps in a single package, with one pair of power connections, yet the power connections are drawn for each opamp, so you can power 1 opamp with a +5V single supply, and another one with +/-7V, all though the same 2 pins. – stevenvh Jul 31 '12 at 10:51
• @stevenvh: I see what you mean. I hadn't noticed that these were all the same package. So the opamp is powered, but the schematic is drawn poorly. – Olin Lathrop Jul 31 '12 at 10:59
• @OlinLathrop, re point 7: I traced out the operation of that circuit and it does seem to take the absolute value of the input. Just to be sure, I checked around on the web and that circuit is here on ESPs site, figure 6A. – Alfred Centauri Jul 31 '12 at 14:57
• @Alfred: OK, I don't have time to think about it carefully so I'll assume for now it is correct. I have removed the original point 7. – Olin Lathrop Jul 31 '12 at 15:22
• Thank you for your comments :) Yes, it takes the absolute value of the input. I am using Altium and I don't know how to put a separate component for power supply as you mention, I will investigate, thanks for the tip. Yes, I forgot decoupling capacitor also. Vcc is 7V and Vee is -7V my input voltage will be no more than 3.5V, it is limited with a voltage dividier. 220 Ω was the value I calculated for the zener, but you are right maybe is too small, as far as I know zener diodes have a minimum current to activate. You are right about D13 and D17, how can I limit output voltage to 3.3V ? Zener ? – Andres Jul 31 '12 at 16:51