Hi,
The analog inputs of the professional driver processors need, ideally, a balanced maximal signal of nearly 8 Vrms.
Here is a circuit, inspired by the Baxandall volume control circuit made popular by Douglas Self, which I built with NE5532s and which apparently fulfills these requirements.
The real nice thing with this kind of circuits relies in the fact is that they use a linear potentiometer to get a log response and that their output are intrinsically of low impedance.
In Baxandall-Self's circuit, the pot cursor is buffered by a voltage follower stage which feeds an inverting voltage gain stage.
In the present circuit, the voltage gain is provided by the first stage, and the second stage only reverts its polarity.
So the outputs of the op-amps deliver the same absolute value of voltage but are of opposite polarity, so they are balanced if the series resistors which connect them to the output world are equal.
Distortion may slightly higher than the Self's circuit.
May be the resistor values shown could be slightly higher to reduce the demands in current of the op-amps. A better solution would be to parallel op-amps as shown is Self's "Small Signal Audio Signal" in the aim to get the lowest noise as possible.
Another feature of the circuit is that the output of the first stage can be used as an unbalanced line without bothering of the inverted polarity of the second stage.
Any impairment with this approach ?
The analog inputs of the professional driver processors need, ideally, a balanced maximal signal of nearly 8 Vrms.
Here is a circuit, inspired by the Baxandall volume control circuit made popular by Douglas Self, which I built with NE5532s and which apparently fulfills these requirements.
The real nice thing with this kind of circuits relies in the fact is that they use a linear potentiometer to get a log response and that their output are intrinsically of low impedance.
In Baxandall-Self's circuit, the pot cursor is buffered by a voltage follower stage which feeds an inverting voltage gain stage.
In the present circuit, the voltage gain is provided by the first stage, and the second stage only reverts its polarity.
So the outputs of the op-amps deliver the same absolute value of voltage but are of opposite polarity, so they are balanced if the series resistors which connect them to the output world are equal.
Distortion may slightly higher than the Self's circuit.
May be the resistor values shown could be slightly higher to reduce the demands in current of the op-amps. A better solution would be to parallel op-amps as shown is Self's "Small Signal Audio Signal" in the aim to get the lowest noise as possible.
Another feature of the circuit is that the output of the first stage can be used as an unbalanced line without bothering of the inverted polarity of the second stage.
Any impairment with this approach ?
Attachments
Last edited:
Hi, nice idea ! Out of curiosity i simmed it, & THD @ 1kHz was 0.001%. I always thought that the original idea published in, i think WW back in the day, seemed a neat solution. I had a feeling it was JLH that did it though ?
The only "slight" negative to it, is the extra time delay between the + & - outputs. Even though it's very small, & probably won't concern many people, it's worth noting.
Thanx for posting
The only "slight" negative to it, is the extra time delay between the + & - outputs. Even though it's very small, & probably won't concern many people, it's worth noting.
Thanx for posting
Hi, nice idea ! Out of curiosity i simmed it, & THD @ 1kHz was 0.001%. I always thought that the original idea published in, i think WW back in the day, seemed a neat solution. I had a feeling it was JLH that did it though ?
Thanx for posting
Hi zeroD,
As far as I know the initial idea of an active log volume control using linear pots was from Peter Baxandall and published in "Audio Gain Controls" part 2,Wireless World article, November 1980 (basic, circuit page #80).
It is quite curious that, apart from Douglas Self, it is so rarely used,
I can scan the two-parts article for those interested.
The only thing I changed in my circuit is to provide some voltage gain in the non-inverting first stage and to use the second stage just to invert the signal.
I quickly measured the harmonic distortion at 1 kHz, it is consistent with what you can expect from an NE5532. I foresee to double (or more) every op-amp to get lower noise and enhance the availability of current.
I have doubts about a delay between the + and - signals at the balanced output .
It would happen with all active circuits providing a balancing feature.
There has even been a further development, where a single linear pot controls two channels simultaneously, also in E&WW, IIRC.
BTW, nice extension of the Baxandall circuit
I think I remember that one but had no idea how it was implemented.
I'll have to search in my E&WW collection.
Last edited:
Sure. Baxandall proposed the same scheme in the article I mentioned, fig. #22 and which Bruno may not be aware of.
In fact it isn't in EWW: it is an idea from the great Steve Woodward:
One Single-Section Potentiometer Sets The Gain On Two Channels | Components content from Electronic Design
That one has a linear gain control; maybe with little more ingenuity would it be possible to approximate a log law?
This was done around 15 yrs ago... application note from Maxim I think. It works well in theory, in practice the pot needs to be 100.000% perfect and so to the layout. Even an ohm or two of resistance in the wiper contact ruins the performance.
R5 and R5A are the pot.
Attachments
Then they probably lifted it from Steve: he must be one of the most plundered author in his field....
Here the picture of the circuit. It has been added a buffer, made of four parallel sections of NE5532 (it was planned to have eight sections), to drive 32 Ohm headphones up to 2 Vrms through a 6.3 mm stereo jack. The balanced output sockets are 6.3 mm jacks instead of the XLRs shown on the schematics :
Here is the whole schematics combining volume control, headphones amp and balanced output (10 Vrms max). The buffer for the phone output needed an input attenuator :
I did not find the input attenuator of the buffer very smart.
So I finally revert to the Self's initial configuration, with an input voltage follower, but now, it is made of parallel NE5532s to be able to drive the 32 Ohm headphones. The voltage gain of 2.5 times is obtained with an inverting stage, and an additional unity gain inverting stage is used to obtain a balanced output stage. I've not tested the circuit yet :
An externally hosted image should be here but it was not working when we last tested it.
Here is the whole schematics combining volume control, headphones amp and balanced output (10 Vrms max). The buffer for the phone output needed an input attenuator :
An externally hosted image should be here but it was not working when we last tested it.
I did not find the input attenuator of the buffer very smart.
So I finally revert to the Self's initial configuration, with an input voltage follower, but now, it is made of parallel NE5532s to be able to drive the 32 Ohm headphones. The voltage gain of 2.5 times is obtained with an inverting stage, and an additional unity gain inverting stage is used to obtain a balanced output stage. I've not tested the circuit yet :
An externally hosted image should be here but it was not working when we last tested it.
Hi,
I hope I do not violate any forum rules if I upload here the original Baxandall articles on active gain stages published in October and November 1980 in Wireless World, as they are so difficult to come by.
They can be downloaded on American Radio History: Documenting the History of Radio TV and FM broadcasting but it's a pain, because you have to search each and every scanned page with the search engine. I save you this painstaking work.
I hope I do not violate any forum rules if I upload here the original Baxandall articles on active gain stages published in October and November 1980 in Wireless World, as they are so difficult to come by.
They can be downloaded on American Radio History: Documenting the History of Radio TV and FM broadcasting but it's a pain, because you have to search each and every scanned page with the search engine. I save you this painstaking work.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.