For this application, electric bass pre-amp, battery opperated, I'd go with a lower power opamp such as the TL061 (single) or TL062 (dual).
Each opamp unit drains only 0,2mA. (0,2mA TL061 and 0,4mA TL062)
In addition, I'd use a single 9V battery and decouple both input and output with bipolar capacitors with cutoff frequency of about 1Hz or less.
In audio path you don't want to have any type of DC level, specially on potentiometers.
Single reverse diode on battery input and you'll be safe.
Polarization at half of battery voltage is so simple, just 2 resistors at the opamp input.
Each opamp unit drains only 0,2mA. (0,2mA TL061 and 0,4mA TL062)
In addition, I'd use a single 9V battery and decouple both input and output with bipolar capacitors with cutoff frequency of about 1Hz or less.
In audio path you don't want to have any type of DC level, specially on potentiometers.
Single reverse diode on battery input and you'll be safe.
Polarization at half of battery voltage is so simple, just 2 resistors at the opamp input.
They connect EV batteries in parallel but they all start from zero volts and then charged up to stop massive current flow between the batteries.
In my experience if you get two identical fresh batteries from the same batch, they are at a very similar voltage and can be connected in parallel without significant discharge. (Parallel: negative leads connected and positive leads connected to form a single negative-positive lead pair)
Yes, indeed. That is, usually the current an op-amp draws increases very slightly with voltage, it certainly won't decrease.
Thanks for your input. Together with @MarcelvdG's information about power consumption, it seems a single supply approach is a no brainer really.
Single supply circuits can have a (often very large) start-up thump though, as the input and output coupling capacitors charge up... You can make a symmetric circuit though, with a bit of careful design. Just be aware that there's no such thing as matched pairs of electrolytic capacitors...
@Mark Tillotson by symmetric circuit I presume you mean something like what I initially had in mind? With two batteries and a true ground reference? If you can expand on the important points of careful design for the symmetric circuit you mention, I'd be truly grateful.
A symmetric virtual ground for instance, using symmetric film caps in the midrail divider followed by opamp follower. Then at power up the virtual ground is actually at mid-rail throughout the transient. Main bulk decoupling needs to be between the two rails, not to (virtual) ground...
@Mark Tillotson hello again. Is this somewhat close to the circuit you describe? I'd love to hear your feedback/corrections.
The left hand side is what I expect to generate the virtual ground, but you've shorted it to the Vin... Call the negative side of the battery V-, and the positive side V+, then GND will be the output of U2.
Hello all, happy new year!
@Mark Tillotson I did check the circuit and it seems to work as intended. By connecting the first op amp's output to the non-inverting input of the audio op amp I am just biasing the input signal to the bias voltage. I did have trouble running simulations to verify it works. I discovered that the reason was my sine signal generator needed to be referenced to 4.5V instead of 0V otherwise the simulation would fail. When I set dc to 4.5V, DC operating point simulation ran, and I also did a transient analysis and AC sweep. Here's the final circuit, and related pictures. This is with just buffering, without any gain.
So I think it functions well. The only thing I didn't manage to do is somehow show Vin swing around 0V instead of the bias voltage (without affecting the simulations), but that's no biggie.
@Mark Tillotson I did check the circuit and it seems to work as intended. By connecting the first op amp's output to the non-inverting input of the audio op amp I am just biasing the input signal to the bias voltage. I did have trouble running simulations to verify it works. I discovered that the reason was my sine signal generator needed to be referenced to 4.5V instead of 0V otherwise the simulation would fail. When I set dc to 4.5V, DC operating point simulation ran, and I also did a transient analysis and AC sweep. Here's the final circuit, and related pictures. This is with just buffering, without any gain.
So I think it functions well. The only thing I didn't manage to do is somehow show Vin swing around 0V instead of the bias voltage (without affecting the simulations), but that's no biggie.
Hello all,
I thought about revisiting this thread. In order to cut down on parts/space/cost I am pondering creating a bias voltage with a simple voltage divider and leaving it unbuffered. The objective is to balance low current draw from the voltage divider, but with enough current to maintain a stable bias voltage for the two non inverting inputs of the TL072BCP op-amp I intend to use. So the focus is on the voltage divider.
I will be using a simple 100u/35V electrolytic for bulk decoupling. For each voltage divider resistor, a film capacitor of 0.1u/50V or, if it offers increased stability, 0.22u/50V since these are the two types of film capacitors I have handy. The op amp will consume about 3.8mA (measured at +-9VDC from a symmetric power supply) and I would like to keep the current draw of the voltage divider as low as possible without compromising bias voltage stability. So the main variable I am investigating for is the value of the two resistors R1, R2.
I have made some modifications to the circuit, mainly introducing a bias voltage injecting resistor of 1M, which also sets the input impedance of the op-amp if I understand correctly. Here's what I have currently:
For reference, I have also come across this schematic on another thread. Apparently it is the input buffer stage of a well known guitar pedal:
Here we obviously have a non symmetric electrolytic capacitor to stabilize the bias voltage, and more importantly, 100k voltage divider resistors. But this device is typically powered by an AC/DC adapter whereas I intend to use my buffer onboard an electric bass using two regular 9V alkaline batteries. Using 100k resistors would yield a voltage divider current of about 0.1mA which I feel is significant, about 3% of total current draw. So the question is, how high can I go with R1, R2 while maintaining a stable bias voltage point for this application? I believe this depends on the input bias current of the op amp which according to the datasheet is 65pA typical, 200pA maximum. But I can't make calculations to estimate what resistors would be appropriate. If there's practical ways of checking if I have a stable bias voltage given specific resistor values, I 'd love to hear about them.
Since the buffers will be used with passive electric bass pickups, I thought of giving you an intuitive idea of what the op amp and bias voltage point should have to deal with. That is, low frequencies, but also fast transients and high peaks with certain playing styles. So here is Wojtek Pilichowski for you, to hopefully make this thread a bit more flashy and colorful. 🙂
I thought about revisiting this thread. In order to cut down on parts/space/cost I am pondering creating a bias voltage with a simple voltage divider and leaving it unbuffered. The objective is to balance low current draw from the voltage divider, but with enough current to maintain a stable bias voltage for the two non inverting inputs of the TL072BCP op-amp I intend to use. So the focus is on the voltage divider.
I will be using a simple 100u/35V electrolytic for bulk decoupling. For each voltage divider resistor, a film capacitor of 0.1u/50V or, if it offers increased stability, 0.22u/50V since these are the two types of film capacitors I have handy. The op amp will consume about 3.8mA (measured at +-9VDC from a symmetric power supply) and I would like to keep the current draw of the voltage divider as low as possible without compromising bias voltage stability. So the main variable I am investigating for is the value of the two resistors R1, R2.
I have made some modifications to the circuit, mainly introducing a bias voltage injecting resistor of 1M, which also sets the input impedance of the op-amp if I understand correctly. Here's what I have currently:
For reference, I have also come across this schematic on another thread. Apparently it is the input buffer stage of a well known guitar pedal:
Here we obviously have a non symmetric electrolytic capacitor to stabilize the bias voltage, and more importantly, 100k voltage divider resistors. But this device is typically powered by an AC/DC adapter whereas I intend to use my buffer onboard an electric bass using two regular 9V alkaline batteries. Using 100k resistors would yield a voltage divider current of about 0.1mA which I feel is significant, about 3% of total current draw. So the question is, how high can I go with R1, R2 while maintaining a stable bias voltage point for this application? I believe this depends on the input bias current of the op amp which according to the datasheet is 65pA typical, 200pA maximum. But I can't make calculations to estimate what resistors would be appropriate. If there's practical ways of checking if I have a stable bias voltage given specific resistor values, I 'd love to hear about them.
Since the buffers will be used with passive electric bass pickups, I thought of giving you an intuitive idea of what the op amp and bias voltage point should have to deal with. That is, low frequencies, but also fast transients and high peaks with certain playing styles. So here is Wojtek Pilichowski for you, to hopefully make this thread a bit more flashy and colorful. 🙂
Last edited:
interesting problem and potential solution, i would like to know more about what you see as problems with a passive selector or "blend pot?
it has me thinking about whether or not this could be done with the right impedance matching transformers....???
awesome vid, thanks for that!
it has me thinking about whether or not this could be done with the right impedance matching transformers....???
awesome vid, thanks for that!
@bucks bunny indeed, I have realized my blunder. What we get is double the voltage headroom, not battery life. But the question at the moment is how to size the voltage divider resistors so that they are as high value as possible without compromising bias voltage stability for this specific op amp and capacitor setup.
Re post #32:
Cf_1 injects supply rail noise direct into the signal path - lose it? If you signals are ground referenced then the bias voltage (aka virtual ground) should be coupled only to ground via Cf_2... However due to the rest of the signal path and that 1M resistor this will only affect very low frequencies - in fact its not clear even Cf_2 is doing anything at all.
Frankly I think you can lose Rbias, Cf_1 and Cf_2. The input signal is far lower impedance than the bias.
Is Rsrc1 an attempt to simulate a source impedance of 10k? If so its in the wrong place - it certainly has zero effect on the simulation as its stands as its short-circuited by the voltage source V2.
Cf_1 injects supply rail noise direct into the signal path - lose it? If you signals are ground referenced then the bias voltage (aka virtual ground) should be coupled only to ground via Cf_2... However due to the rest of the signal path and that 1M resistor this will only affect very low frequencies - in fact its not clear even Cf_2 is doing anything at all.
Frankly I think you can lose Rbias, Cf_1 and Cf_2. The input signal is far lower impedance than the bias.
Is Rsrc1 an attempt to simulate a source impedance of 10k? If so its in the wrong place - it certainly has zero effect on the simulation as its stands as its short-circuited by the voltage source V2.
But that's only got 1MHz GBP and 3.5V/us slew rate - not what I'd call adequate for audio. Most audio capable opamps have 2 to 5mA supply current ratings for good reason, they need it...
Considering voltage output full rail swinging from -4 to +4V, the minimum slew rate you need for 20kHz is 0.5V/usec, so 3.5V/usec is enough.
SR = 2*π*f*Vpeak
And considering this is a bass pre-amp, harmonics around 20kHz will be much lower than full rail.
He can build the pre-amp with an 8pin DIP socket to try TL071 and TL061.
Listen to both options and/or measure the results and take the decision.
I think the battery saving for this application is relevant and it's worth the optimization.
The TL061 datasheet shows the maximum voltage swing versus frequency, which confirms 20kHz is ok.
SR = 2*π*f*Vpeak
And considering this is a bass pre-amp, harmonics around 20kHz will be much lower than full rail.
He can build the pre-amp with an 8pin DIP socket to try TL071 and TL061.
Listen to both options and/or measure the results and take the decision.
I think the battery saving for this application is relevant and it's worth the optimization.
The TL061 datasheet shows the maximum voltage swing versus frequency, which confirms 20kHz is ok.
@ron68 thank you very much for your input. I am aware of the tl062, and was in fact planning to evaluate it since its performance may be good enough for electric bass and its low power consumption is very appealing. I do have one installed in the preamp of one of my basses.
If you notice my last schematic, I intend to move to 18V total power supply, two alkaline 9V batteries in series. For this initial build I want more than ample headroom. Since the tl072 swings at 1.5V distance from its rails I would like to completely eliminate the possibility of a very hot output humbucker signal getting clipped or distorted. If 9V proves enough in practice, I will revert to that.
I presume this means wider output voltage swing for your tl061 recommendation, although gain will be set to unity, no amplification. I bet this characteristic curve also applies to the tl072 in an analogous way.
If you notice my last schematic, I intend to move to 18V total power supply, two alkaline 9V batteries in series. For this initial build I want more than ample headroom. Since the tl072 swings at 1.5V distance from its rails I would like to completely eliminate the possibility of a very hot output humbucker signal getting clipped or distorted. If 9V proves enough in practice, I will revert to that.
I presume this means wider output voltage swing for your tl061 recommendation, although gain will be set to unity, no amplification. I bet this characteristic curve also applies to the tl072 in an analogous way.
Last edited:
Hi,
If you move to swing output to +/-7.5V (9-1.5), the minimum slew rate doubles and you'll need at least 0.94V/us at 20kHz, which is still within the op amp capability.
And as we do in power amplifiers, if you want to make sure no high speed signal (not audio) interferes with the op amp slew rate, you can put a low pass filter in the input so as to make sure the bandwidth is limited to 20 or 30kHz.
If you move to swing output to +/-7.5V (9-1.5), the minimum slew rate doubles and you'll need at least 0.94V/us at 20kHz, which is still within the op amp capability.
And as we do in power amplifiers, if you want to make sure no high speed signal (not audio) interferes with the op amp slew rate, you can put a low pass filter in the input so as to make sure the bandwidth is limited to 20 or 30kHz.