# Variable voltage drop

#### mv1618

Joined Feb 24, 2016
7
Hello!
I have a question concerning a circuit i made. It uses an opamp as an voltage reference, which is variable, and can thus control the voltage that appears on the OUT.

The circuit acts like there is a variable voltage drop accross a shunt diode. The voltage drop is equal to Vref and the drop of the diode used. The circuit seems to be wrking fine on Ltspice. However, i need many of those circuits in a thing i'm building and using so many opamps seems to be space consuming since they would be all wired as a buffer. Is there another way to make a voltage reference with very little space consuming parts and still very low output impedance. The current capability doesn't have to be high, but it is important that no series resistance appears to be in series with the diode (max few hundred ohms). Any ideas?

#### dl324

Joined Mar 30, 2015
16,697
How are these voltage references to be used?

If you don't want series resistors, do you actually want current sinks/sources?

#### WBahn

Joined Mar 31, 2012
29,868
What do you mean that there be no series resistance in series with the diode? What is R1 in your schematic if not a resistance that is in series with the diode?

Do you need the same vref for all of the diodes, or just a single Vref that serves all of them?

It would be really helpful if you sketched a simplified version of what you are trying to accomplish (even if it just has two diodes in it) and explain what the goal is and what is important.

#### mv1618

Joined Feb 24, 2016
7
It is used for precise clipping of a signal - i want the signal to clip at a precise voltage and with adding series resistance i could afffect the gain of the signal when the diode starts conducting.

#### WBahn

Joined Mar 31, 2012
29,868
It is used for precise clipping of a signal - i want the signal to clip at a precise voltage and with adding series resistance i could afffect the gain of the signal when the diode starts conducting.
All the more important to give a better idea of the actual circuit you have in mind. When the diode "starts conducting" it is going to go from a very high resistance, to an intermediate resistance, to a pretty low resistance. How will that affect this "gain" you are concerned about?

#### mv1618

Joined Feb 24, 2016
7
Imagine that the diode forms a potential divider with R. When diode is not conducting, the gain is near 1, and when diode is conducting the output is maintained at more or less constant voltage. That is, if we don't put a resistor in series with the diode (between the diode and the out terminal). By varying the Vref the voltage at which the output stands when the diode is conducting is varied. That's the voltage i want to vary, preferebly without using an opamp.

#### ian field

Joined Oct 27, 2012
6,536
Hello!
I have a question concerning a circuit i made. It uses an opamp as an voltage reference, which is variable, and can thus control the voltage that appears on the OUT.

The circuit acts like there is a variable voltage drop accross a shunt diode. The voltage drop is equal to Vref and the drop of the diode used. The circuit seems to be wrking fine on Ltspice. However, i need many of those circuits in a thing i'm building and using so many opamps seems to be space consuming since they would be all wired as a buffer. Is there another way to make a voltage reference with very little space consuming parts and still very low output impedance. The current capability doesn't have to be high, but it is important that no series resistance appears to be in series with the diode (max few hundred ohms). Any ideas?
The TL431 "programmable zener" is actually a comparator with a built in 2.5V reference. It is available in a compact TO92 package and you need 2 external resistors to set up the nfb and programme its voltage setting.

If you stick with op-amps - you can minimise the pin count by using quads like the LM324 etc.

#### WBahn

Joined Mar 31, 2012
29,868
Imagine that the diode forms a potential divider with R. When diode is not conducting, the gain is near 1, and when diode is conducting the output is maintained at more or less constant voltage. That is, if we don't put a resistor in series with the diode (between the diode and the out terminal). By varying the Vref the voltage at which the output stands when the diode is conducting is varied. That's the voltage i want to vary, preferebly without using an opamp.
Again, do you need to be able to vary the clipping voltage independently, or will it always be the same clipping voltage for all channels?

How constant does the voltage need to be?

How much current will each channel need to absorb?

#### AnalogKid

Joined Aug 1, 2013
10,944
How many circuits?
What is the voltage range of IN?
What is the voltage range of Vref?
What is the value of R1?
What is the required accuracy of the clipping voltage level at Out?
What is the amount of overvoltage, the difference between IN and (Vref + Vdiode)? Note that as the input voltage increases above the clip level, Vdiode will increase, so the clipping level will have some of the original signal amplitude modulated on top of the DC clip voltage.

The answers combine to tell us the amount of current and headroom an alternate circuit would need.

ak

#### ronv

Joined Nov 12, 2008
3,770
It is used for precise clipping of a signal - i want the signal to clip at a precise voltage and with adding series resistance i could afffect the gain of the signal when the diode starts conducting.
Do you want them all to clip at the same level or do you have many Vrefs?
I need to refresh more often.

#### Alec_t

Joined Sep 17, 2013
14,224
The diode voltage drop is also temperature and current dependent. How important is that?

#### mv1618

Joined Feb 24, 2016
7
I need to vary the voltage indipendently. The voltage doesn't need to be too constant (a few mV up or down). It will be small currents, less than 1mA. It is about 8 of those circuits, voltage range will probably be 18V, value of R1 in tens of Kohms, required accuracy is about +/-20mV, overvoltage will vary, but max 9V. Temperature and current dependency is neglible. TL431 gives me weird results in spice, as it apparently needs to stabilise, resulting in a spike on the waveform, thus unusable for AC, I'm guessing. The most important thing is the ratio between voltages that multiple circuits will be at.

#### AnalogKid

Joined Aug 1, 2013
10,944
So, rather than one opamp driving 8 diodes and resistors, you need 8 independently adjustable references? yes/no

You should be able to eliminate the diode and its errors, and clip the output directly by using one section of an open collector output LM339. Haven't tried it, but it looks good on paper.

ak

#### mv1618

Joined Feb 24, 2016
7
I don't just need indipendently adjustable references, i need the variable clipping characteristis that diodes and resistors offer.
I won't use the LM339 as the soft knee diodes have is desirable.

#### ronv

Joined Nov 12, 2008
3,770
These two statements don't make sense to me. Maybe you can draw two circuits and show the voltages on both.

I need to vary the voltage indipendently.
I don't just need indipendently adjustable references

#### mv1618

Joined Feb 24, 2016
7
Nevermind the second one, i was wanting to tell that i need more than just that for my purpose. Is there a thing like an IC that would contain just unity gain buffers to make it spae economical?

#### g6ypk

Joined Nov 15, 2013
2
Nevermind the second one, i was wanting to tell that i need more than just that for my purpose. Is there a thing like an IC that would contain just unity gain buffers to make it spae economical?
You need to go back to basics. You clearly have no background in design. I am not trying to be unkind, but you need to rethink your capabilities.

#### dl324

Joined Mar 30, 2015
16,697
Is there a thing like an IC that would contain just unity gain buffers to make it spae economical?
Yes (e.g. LM302, LM310), but you could use quad opamps and get twice as many in a slightly smaller footprint as the hardwired voltage followers.

#### AnalogKid

Joined Aug 1, 2013
10,944
If you need 18 V at the diode cathode, and each circuit independently adjustable, I don't know what else there is beyond a quad opamp running on 21+ V.

If you need just a bit less and all the same, enter CMOS. A CMOS hex inverter has 6 gates in 14 pins. If you ground all of the inputs, then varying Vdd to the chip varies the output voltage at all 6 outputs.

ak