biasing this op amp

[AFex54]

I was hoping i could finally move on to the final wiring of my project but i realised I had forgotten a vital piece.
I needed a very simple sample and hold circuit and this is my design so far:

It sample perfectly but Im struggling to get the op amps working the way they should,
they are 1.5V at the outputs with 0V at the inputs
I know I need a virtual ground but I dont know how to implement it in this case,
any ideas?

 

[ver chan]

where did you get this schematic?

 

[Harald Kapp]

The image doesn't show here (probably blocked by firewall). Please upload the image to this forum (upload button on bottom right of edit window) and link to it in your post ("Image" button in upper menu row).

 

[LvW]

You have forgotten any common ground reference!

 

[AFex54]

You have forgotten any common ground reference!

I know, that what the part i need help with!
im using a split supply and tried connecting the virtual ground in many different ways but ever time the circuit stops working when i do

The image doesn't show here (probably blocked by firewall). Please upload the image to this forum (upload button on bottom right of edit window) and link to it in your post ("Image" button in upper menu row).

oh ok didnt realise you could upload here

where did you get this schematic?

I made it

 

[Harald Kapp]

You need either a virtual ground as can be generated in many ways.
Another solution is using rail-to-rail operational amplifiers (or at least ones that go all the way to 0V). These can go to 0V without using a split supply.

 

[LvW]

You have mentioned to use split supply. In this case, you do NOT need a virtual ground node.
Instead, use the pos. supply voltage between the corresponding opamp pin and ground - and do the same with the negative supply voltage.
In this case, the ouput voltages of the opamps will be centered at 0 volts (ground).

 

[hevans1944]

It would be a big help if you would label the pin numbers and the type of op-amp you are using.

If that 9 V battery is intended to be a "split supply" then you need two of them: one to provide +9 V to the positive supply rail of the op-amp and one to provide -9 V to the negative supply rail of the op-amp.

However, the circuit "as drawn" should work, but probably not as you expect. With the potentiometer at its mid-point applying, +4.5 V to the non-inverting input of the first op-amp with respect to the negative terminal of the battery, the output of the first op-amp should be +4.5 V with respect to the negative terminal of the 9 V battery. The same applies to the second op-amp after the switch is closed to charge the sampling capacitor: its output should be +4.5 V with respect to the negative terminal of the 9 V battery.

As you move the wiper arm toward the positive end of the battery, the output should increase above +4.5 V with respect to the negative terminal of the battery. Similarly, as you move the wiper arm toward the negative end of the battery, the output should decrease below +4.5 V with respect to the negative terminal of the battery. Here's the rub: most op-amps cannot swing their outputs rail-to-rail, meaning the output cannot go fully positive to 9 V with respect to the negative terminal of the battery; nor can the output go fully to zero with respect to the negative terminal of the battery. If you want the output swing to include zero, then you need a negative supply rail for the op-amp.

This is most easily accomplished by adding another battery in series with the first one and designating their common connection as "ground" for your circuit. Your output would then be measured with respect to this "ground" connection. The 100 kΩ potentiometer remains connected between the positive and negative supply rails of the op-amps, but now it causes the outputs to swing both positive and negative with respect to your new "ground" connection.

The link that @Harald Kapp provided should be read for alternative means of providing a "virtual ground" connection if you are unable or unwilling to add a second battery to your circuit.

 

[AFex54]

It would be a big help if you would label the pin numbers and the type of op-amp you are using.

If that 9 V battery is intended to be a "split supply" then you need two of them: one to provide +9 V to the positive supply rail of the op-amp and one to provide -9 V to the negative supply rail of the op-amp.

However, the circuit "as drawn" should work, but probably not as you expect. With the potentiometer at its mid-point applying, +4.5 V to the non-inverting input of the first op-amp with respect to the negative terminal of the battery, the output of the first op-amp should be +4.5 V with respect to the negative terminal of the 9 V battery. The same applies to the second op-amp after the switch is closed to charge the sampling capacitor: its output should be +4.5 V with respect to the negative terminal of the 9 V battery.

As you move the wiper arm toward the positive end of the battery, the output should increase above +4.5 V with respect to the negative terminal of the battery. Similarly, as you move the wiper arm toward the negative end of the battery, the output should decrease below +4.5 V with respect to the negative terminal of the battery. Here's the rub: most op-amps cannot swing their outputs rail-to-rail, meaning the output cannot go fully positive to 9 V with respect to the negative terminal of the battery; nor can the output go fully to zero with respect to the negative terminal of the battery. If you want the output swing to include zero, then you need a negative supply rail for the op-amp.

This is most easily accomplished by adding another battery in series with the first one and designating their common connection as "ground" for your circuit. Your output would then be measured with respect to this "ground" connection. The 100 kΩ potentiometer remains connected between the positive and negative supply rails of the op-amps, but now it causes the outputs to swing both positive and negative with respect to your new "ground" connection.

The link that @Harald Kapp provided should be read for alternative means of providing a "virtual ground" connection if you are unable or unwilling to add a second battery to your circuit.

You have mentioned to use split supply. In this case, you do NOT need a virtual ground node.
Instead, use the pos. supply voltage between the corresponding opamp pin and ground - and do the same with the negative supply voltage.
In this case, the ouput voltages of the opamps will be centered at 0 volts (ground).

sorry, I meant to say I have a single supply,
I created a virtual ground with a 4.7k resistor divider,
but im strugglin to communicate it with the circuit,
i tried connecting to the noninverting input but it doesnt work
Also I am using a TL082 , would I need to get a rail to rail op amp?

 

[hevans1944]

So, you connected two 4.7 kΩ resistors in series across your 9 V battery, and are using their common junction as your "virtual ground?" If so, that is what you measure your output with respect to, the junction of those two resistors. You should be able to get the TL082 outputs to swing ±3 V or so with respect to that "virtual ground" using a 9 V battery for the supply.

You should also "bypass" the two 4.7 kΩ resistors with a pair of 220 μF electrolytic capacitors.

Connect the bottom of your "sampling capacitor" to your "virtual ground".

Measure the "held output, 0 - 9V" with respect to the "virtual ground" but realize it will now be ±3 V instead of zero to +9 V.

But, as I said earlier, your circuit should work "as is" if you measure the output with respect to the negative terminal of the battery. Just don't expect it to swing more than about +1.5 V to +7.5 V, NOT 0 to 9 V. Refer to the TL082 datasheet for output swing versus supply voltage limitations.

 

[AFex54]

You need either a virtual ground as can be generated in many ways.
Another solution is using rail-to-rail operational amplifiers (or at least ones that go all the way to 0V). These can go to 0V without using a split supply.

So, you connected two 4.7 kΩ resistors in series across your 9 V battery, and are using their common junction as your "virtual ground?" If so, that is what you measure your output with respect to, the junction of those two resistors. You should be able to get the TL082 outputs to swing ±3 V or so with respect to that "virtual ground" using a 9 V battery for the supply. You should also "bypass" the two 4.7 kΩ resistors with a pair of 220 μF electrolytic capacitors.

the trouble is I need to be able to sample the full 9V, is this out of the question?
would there be a suitable rail to rail ,single supply chip I could get?

 

[hevans1944]

the trouble is I need to be able to sample the full 9V, is this out of the question?
would there be a suitable rail to rail ,single supply chip I could get?

Of course it out of the question with the circuit and components you have. It is what it is. You could try to substitute an op-amp with rail-to-rail output capability. Google is your friend. Maybe the attached file will help.

 

[AFex54]

Of course it out of the question with the circuit and components you have. It is what it is. You could try to substitute an op-amp with rail-to-rail output capability. Google is your friend. Maybe the attached file will help.

I meant if it was out of the question with a single supply, the components dont matter as long it does what i need.
thanks, the OPA4342 seems like it would work.
the only catch with rail-to-rails is the bigger price tag then?

 

[AFex54]

http://ie.rs-online.com/web/p/operational-amplifiers/6611577/
this is what i could find on RS,
seems to be fit for the purpose, operates with 9V single supply ,rail to rail,
but just want to make sure im not overlooking anything since their hardly cheap

 

[hevans1944]

http://ie.rs-online.com/web/p/operational-amplifiers/6611577/
this is what i could find on RS,
seems to be fit for the purpose, operates with 9V single supply ,rail to rail,
but just want to make sure im not overlooking anything since their hardly cheap

Yikes! I would certainly investigate another solution at that price and minimum order quantity.

Perhaps we could help you find other solutions if you would tell us what you are trying to do, rather than you providing your proposal of how to do it. There is usually more than one solution to any engineering problem. Answers to the following questions can possibly yield other design solutions: What is the purpose of the 100 kΩ potentiometer? Why is the power supply 9 V DC? What kind of switch are you using ahead of the sampling capacitor? What is the output used for? What are the accuracy requirements? How much droop, drift, and overshoot can be tolerated in the sample-and-hold? What is the sampling interval? How long is the switch closed to acquire a sample?

Here is an additional selection of op-amps that may work.

 

[AFex54]

Yikes! I would certainly investigate another solution at that price and minimum order quantity.

Perhaps we could help you find other solutions if you would tell us what you are trying to do, rather than you providing your proposal of how to do it. There is usually more than one solution to any engineering problem. Answers to the following questions can possibly yield other design solutions: What is the purpose of the 100 kΩ potentiometer? Why is the power supply 9 V DC? What kind of switch are you using ahead of the sampling capacitor? What is the output used for? What are the accuracy requirements? How much droop, drift, and overshoot can be tolerated in the sample-and-hold? What is the sampling interval? How long is the switch closed to acquire a sample?

Here is an additional selection of op-amps that may work.

yes prices really are terrible but the delivery is extremely fast and very cheap.
unfortunately the circuit I drew doesnt have any room for change other than the sample and hold itself.
its gets rather complicated if I try to explain it but here goes
9V is the supply of the synthesizer ive built, the 100K pot is voltage divider used to control the frequency of an oscillator, there is one for every key on my CV keyboard
the switch is a JFET and which im gating with 9V from a manual switch e.g the gate from my CV keyboard .
the sample and holds purpose is to maintain the control voltage from the potentiometer of any given key after you let go of the key to accommodate the release of the envelope generator,
the switch will be closed during the sustain of a note.
Im using polystyrene caps which will compensate for some droop , but I have tested with just a ceramic cap within the 1.5V-7.5v range of the TL082 and the droop is still within reasonable levels.
the oscillator is 18mV per octave so there could be issues if the sampleing isnt smooth.
aslong as any offset from the pot output voltage isnt huge (no bigger than 25mV if had to estimate) and aslong as it stays consistent it wont matter.
Anyway i will just buy the previous op amp I posted,
most of the MFOS sample and hold circuits for proper +/-12V musically accurate modular synths are still super simple apart from the polystyrene caps.

 

[CDRIVE]

Since your original posting of this here...
https://www.electronicspoint.com/threads/will-a-basic-buffer-be-sufficient.274460/
I've spent some time studying the Noise Toaster and reading the plethora of data that the author (Ray Wilson) provided on his website. I've also watched two of his videos which are quite articulate.

I have no doubt that we could eventually supply an acceptable solution as per your requirements but I suspect that Mr. Wilson could accomplish this far more expeditiously.

FYI: There's a good reason why Mr. Wilson chose the 324 OpAmp in his noise toaster. While they won't swing up to the + rail they do work well with single ended power supplies as he's employing them. Zero offset is quite low.

Chris

 

[AFex54]

Since your original posting of this here...
https://www.electronicspoint.com/threads/will-a-basic-buffer-be-sufficient.274460/
I've spent some time studying the Noise Toaster and reading the plethora of data that the author (Ray Wilson) provided on his website. I've also watched two of his videos which are quite articulate.

I have no doubt that we could eventually supply an acceptable solution as per your requirements but I suspect that Mr. Wilson could accomplish this far more expeditiously.

FYI: There's a good reason why Mr. Wilson chose the 324 OpAmp in his noise toaster. While they won't swing up to the + rail they do work well with single ended power supplies as he's employing them. Zero offset is quite low.

ChrisView attachment 20744

ok, I sent Ray an email
It was a lot of info for one email but I think hes familar enough with the designs to understand it
Also 0-8 would be a sufficient range aslong as it 1-9V,
the reason being is the noise toaster oscilator is capacble of very low frequencies and it becomes inaudable at around 1V and below
edit: or is it lower voltage/higher frequency? if so then im good to go with 0-8V

 

[hevans1944]

Ah hah! An old-school analog designer! With a musical inclination too. I love it. Well, sometimes the choices we make up front have long-term unforeseen consequences down stream... but that doesn't mean the choices were wrong. Perhaps just unfortunate or inconvenient with the benefit of 20-20 hindsight. I have to assume the keyboard is chromatically tuned by all those potentiometers (one for each key), which is a job I would not have undertaken for more than one octave of notes... maybe not even for that many.

Many years ago I owned a Conn electronic organ, a low-end consumer model probably sold more as furniture than as a musical instrument. It used individual vacuum tube oscillators for each key, twelve per octave, and wave-shaping circuits switched with hundreds of little neon lamps mounted on the keyboard switches to implement the voices for the stops. It featured quite a range of voices with two manuals and an octave of foot keys, plus a rotating Leslie speaker and sufficient audio power to fill a small living room with... um... music.

I never did learn enough music to play the thing well, but it did serve to hook me on electronic music for awhile. I got involved with MIDI and synthesizers and computers and yada yada yada. I hooked up with a few musicians and we explored things electronically. It was a fascinating era (the 1970s) as old-school analog techniques made room for "modern" digital techniques. I was amazed to discover just how sophisticated those analog techniques had become and how quickly digital synths appeared to be taking over. But, at least for me, MIDI and electronic music was just a passing fancy. Plus the industry was moving too fast for me to keep up. My expensive Roland synth became obsolete as MS-DOS was replaced by Windoze. My midi piano somehow lost its metal fold-up stand and I got tired of trying to learn how to play it on my lap. My stereo "hi-fi" rig became less useful as it was relegated to the basement. I think the final straw occurred when the cat pissed on my Jensen speakers, so I had to abandon them to the dump. Things were never the same after that.

For awhile I hooked up with a friend who was fascinated with pipe organs and we visited several other people with similar interests. It turns out you need a lot of money to pursue that hobby. One I remember vividly had restored a house in an "historical district" of East Dayton. It featured a detached building in the back yard devoted to model railroading in one large room with the rest of the building containing a rather large pipe organ. This "special interest group" did organ concerts there and other places on selected weekends, by invitation only. It was there that I learned that the "pipes" you see in church organs are fakes: the real noise is produced "back stage" with complicated air handling systems and electrically operated valves. My friend eventually purchased a pipe organ from a church that was being demolished in downtown Dayton to make room for a parking lot. We parted ways soon after that as I didn't share quite that much passion for his hobby. But I did learn the rudiments of "electronic music" during a brief span of five or ten years.

The main take-away I got from all of this? You need to use inexpensive parts to make music because you will need a LOT of them. I don't think there is anything that I can add to your basic design. but if I were starting from scratch I would definitely choose a bi-polar power supply, and perhaps a commercial switch instead of discrete JFETs, just to make things easier to wire up. You will note that your Music From Outer Space circuit does implement a "virtual ground" and uses it where necessary. The biggest problem with sample-and-hold circuits is accounting for op-amp bias current, which causes the output to drift. This can be offset, but that circuit's adjustment is prone to drift with changes in temperature. Best bet is to choose an op-amp with well-matched and very small bias currents. The other alternative is to digitize that analog signal and store the value in a register. Zero drift doing that. If you need the stored value as an analog signal further down the chain, a digital-to-analog converter followed by a low-pass filter recovers it.

Anyhoo, I'm gonna let Chris (@CDRIVE) take it from here and get back to playing with PICs and Arduinos.

Hop

 

[CDRIVE]

ok, I sent Ray an email
It was a lot of info for one email but I think hes familar enough with the designs to understand it
Also 0-8 would be a sufficient range aslong as it 1-9V,
the reason being is the noise toaster oscilator is capacble of very low frequencies and it becomes inaudable at around 1V and below
edit: or is it lower voltage/higher frequency? if so then im good to go with 0-8V

You can check that with a DVM.

Chris