THAT preamp with 3 Band fixed EQ

I want to make a preamp with a simple 3 band eq that can do broad strokes on the low, mids, and highs. I essentially took the common layout for a THAT1510 preamp and added three Sontec 250 EQ bands. I am wondering if anyone could verify the circuit design. Can I connect them together like the schematic link below? I have simulated the EQ part after the DC servo and it appears to be working well. I have also built and tested the preamp part and that works well too. But will they work together...

I do feel that I might have forgotten something. I did remove quite a few elements from the original EQ design, but the simulation is functioning great...

Be sure to click the schematic tab. Ignore the layout tab for now...
https://easyeda.com/editor#id=|6074ad256f4b4691b52041ef9e135004|605eb1a838984dc3882d61975dc0a36b

Post a schematic in an easily viewable form,  like a jpg,  gif,  etc.

If it simulates ok it should probably work ok,  provided you are within normal working limits of the devices.

Here is a picture for easy viewing.

Looks fine to me but I didn't really study the EQ controls. It would take me a week to fully understand what's going on there. I'll assume that whomever designed it, knew what they were doing.

But I don't know about the servo. That can only add more noise compared to just connecting REF to ground. If there really is some reason to be concerned with DC offset, I would just use a coupling cap. Just having the EQ in will distort the signal far more than any coupling cap.

eddie_ruff_ said:

I want to make a preamp with a simple 3 band eq that can do broad strokes on the low, mids, and highs.

Click to expand...

  Yet you have implemented a fully-featured parametric, with Frequency, boosr/cut and BW controls...

squarewave said:

Looks fine to me but I didn't really study the EQ controls. It would take me a week to fully understand what's going on there. I'll assume that whomever designed it, knew what they were doing.

Click to expand...

The Sontec is similar to a GML8200, based on the bridged-T topology.

squarewave said:

But I don't know about the servo. That can only add more noise compared to just connecting REF to ground. If there really is some reason to be concerned with DC offset, I would just use a coupling cap. Just having the EQ in will distort the signal far more than any coupling cap.

Click to expand...

This is the exact area that I have concerns. I think both designs separately had some type of DC servo, so that is why I included it. But setting REF to ground and using a coupling cap might be a great alternative...

abbey road d enfer said:

  Yet you have implemented a fully-featured parametric, with Frequency, boosr/cut and BW controls...

Click to expand...

It may look that way on the schematic since I kept nearly all of the Sontec functionality, but I plan to have the Q control as a trim pot on the PCB (a set and forget feature). I do want boost/cut for every band.

And the frequencies have been fixed as resistors. The boost/cut pot has a push/pull switch for each band which selects the different frequencies. For example, boost/cut 800Hz, then you pull and you have boost/cut 400Hz. So in total, you have boost/cut on 6 different curves but you can only use three at a time. Still a three knob eq...which is pretty simple 

eddie_ruff_ said:

It may look that way on the schematic since I kept nearly all of the Sontec functionality, but I plan to have the Q control as a trim pot on the PCB (a set and forget feature). I do want boost/cut for every band.

And the frequencies have been fixed as resistors. The boost/cut pot has a push/pull switch for each band which selects the different frequencies. For example, boost/cut 800Hz, then you pull and you have boost/cut 400Hz. So in total, you have boost/cut on 6 different curves but you can only use three at a time. Still a three knob eq...which is pretty simple 

Click to expand...

There are other topologies that allow that with less parts and are less challenging, in particular regarding the noise performance.

abbey road d enfer said:

There are other topologies that allow that with less parts and are less challenging, in particular regarding the noise performance.

Click to expand...

Interesting, could you shed some light on these other topologies...any keywords I should be searching for

eddie_ruff_ said:

This is the exact area that I have concerns. I think both designs separately had some type of DC servo, so that is why I included it. But setting REF to ground and using a coupling cap might be a great alternative...

Click to expand...

The only serious justification for a servo is if you drive a transformer.

eddie_ruff_ said:

Interesting, could you shed some light on these other topologies...any keywords I should be searching for

Click to expand...

Just look at the EQ in many Soundcraft mixers, 1600, 2400, Ghost, 6000...

A 3-band broad-stroke EQ is one op-amp, not a whole heap.
https://www.electroschematics.com/6201/3-band-equalizer/

PRR said:

A 3-band broad-stroke EQ is one op-amp, not a whole heap.
https://www.electroschematics.com/6201/3-band-equalizer/

Click to expand...

I have found that putting 3 bands of Baxandall on one op amp leads to too much interaction between the bands (IMO). Splitting it up across two op amps eliminates that interaction, and also inverts the polarity back to correct polarity.

JR

...for those interested, I've finished this build some time ago. I now have 10 channels of preamp and EQ. The low, mid, and high bands have two frequency options (via push-pull). This is an old picture before I put the knobs on. Overall, it sounds great on drums and guitars. Not the simplest of designs, but a breeze to track with and very musical to use.

Sweet. Nice build.

PRR said:

A 3-band broad-stroke EQ is one op-amp, not a whole heap.
https://www.electroschematics.com/6201/3-band-equalizer/

Click to expand...

Very interesting circuit.  Might have to take a whack at that when I get a chance.  I do have one stupid question, though.  There are a couple places in that schematic where they have two 1.8K resistors in series with each other.  Why not just a single 3.6K? 

CurtZHP said:

Very interesting circuit.

Click to expand...

That was commonplace in HiFi equipment. Too cheap to be really good, as JR mentioned. Don't waste time on this, if you want to learn, make the midrange section separate in order to make it slightly better. But it's very broad.

  There are a couple places in that schematic where they have two 1.8K resistors in series with each other.  Why not just a single 3.6K?

Click to expand...

Beware these cookbook circuits have probably never been built by their authors. These resistors define the amount of Boost/Cut. It was typical to contain the range to +/-6dB; more would be ridiculous.

abbey road d enfer said:

That was commonplace in HiFi equipment. Too cheap to be really good, as JR mentioned. Don't waste time on this, if you want to learn, make the midrange section separate in order to make it slightly better. But it's very broad.
Beware these cookbook circuits have probably never been built by their authors. These resistors define the amount of Boost/Cut. It was typical to contain the range to +/-6dB; more would be ridiculous.

Click to expand...

Understood.
Not really looking for surgical EQ, just "a little dab'll do ya!"  That's why this initially got my attention.  When you say "too cheap," do you mean the type of components or lack thereof? 

CurtZHP said:

Understood.
Not really looking for surgical EQ, just "a little dab'll do ya!"  That's why this initially got my attention.  When you say "too cheap," do you mean the type of components or lack thereof?

Click to expand...

The components are OK, but using one opamp for mediocre results when using two would yield much better results is cheap. You may have noticed I'm not a blind proponent of minimalism. As a bonus, putting these two stages in series will result in no phase inversion, which may or may not be an issue.