1-band parametric EQ

[kothoma]

Hey, thanks for the file and the detailed explanations.

So I open the file, start Simulate/Run, probe the output and get two curves. I guess one is the fequency response and the other one is the phase? The third (high shelf) section is fully ccw (dialed to the + side of the opamp) on the R15/R20 pot, that is maximum cut. The other two are neutral. So the solid line is the frequency response and shows a treble cut, right?

The similarity/duality to a gyrator (and low shelf) is stunning. Maybe you can't make it switchable, but the board could be designed to choose either version. So with two sections you could build a two-band parametric eq, or a bass/treble tone control with sweepable working points, or any mixture of these. Or you only polulate one section. Nice. Now is the question if the three lower opamps can be replaced by transistors to support the flexibility.

Hm, how do I change values of components? [right click]

[RobA]

Yes the second line is the phase response and you've got what the controls are setup like correct as well.

Where you run into trouble with the transistor version is the emitter-ground connection. This usually goes through a smallish (1k-to-10k or so) resistor. The problem is that this connection also hits in the feedback path of the buffer through the 470Ω resistor. That causes a problem because you then have the path from ground to the inverting op-amp input of the main op-amp and that screws with the Vref. Maybe you could put a large value cap of say 22µF between the emitter resistor and ground. It'd alter the frequency response some, but maybe not enough to matter.

The selection of shelf versus PEQ can be made with jumpers. That just leave the option of not including a section to figure out. The op-amps are already setup as unity gain buffers, so if a section is not going to be used, we just need to connect the non-inverting input to ground and then it's safe to leave off that band. I think that should make it so that it can be done with a dual op-amp and no real downsides.

The low shelf/PEQ selection could be put on as a toggle. If the toggle is left off, it defaults to PEQ. To select a fixed low shelf, you just leave off the toggle and jumper the cap.

Another possible option is to jumper the frequency pots or resonance pots of either section.

It should make for a pretty flexible little board -- if it can be made small enough. Using SMD op-amps might help, but it's probably not going to get a whole lot of space savings. If these ideas sound good, I can do a layout and see how small it can be. 

[kothoma]

I think you've sorted out the switching/jumping options allright.

So two dual op amps then instead of one quad? More flexibility in the layout?

Now I'm really curious how small this can be made! And I like the 2-for-1 deal.

[RobA]

...
So two dual op amps then instead of one quad? More flexibility in the layout?

Good question. I tend to like duals over quads because of more options available for types of op-amps and selecting for purpose. It also does help in doing some layouts. In this case though, I can't think of any reason to not use the same op-amp in each position and since it'll all be off board pots, there probably won't be much help from using duals for the layout.

I'll try it with a quad and PTH first and see what I can do with that.

But, I'm going to do a bit of breadboarding first. I'll let you know if that verifies the concepts.

[Thomas_H]

Not sure if you want all the fun for yourself, so I added a spoiler :-)

Only the  boost/attenuate and the volume pot are outside. Q and Freq adjustment are inside via trimmers.


SPOILER
-----------------------------------
Schematic & Layout:
http://diy.thcustom.com/wp-content/uploads/2014/09/1-band-parameq_V1.0.pdf

One of the prototypes:

[kothoma]

No spoiler at all, Thomas. Nice to see how it could be done. Using the fourth opamp as a booster can be useful in many situations. Your layout shows that there's not much space to waste.

Our litte discussion here has slowly drifted towards a 2-band parametric EQ. Will you make your board available in your shop? It could come in handy.

[kothoma]

But, I'm going to do a bit of breadboarding first. I'll let you know if that verifies the concepts.

Thanks, and take your time. Meanwhile I'm having great fun playing with values in your LTspice model!

[Thomas_H]

No spoiler at all, Thomas. Nice to see how it could be done. Using the fourth opamp as a booster can be useful in many situations. Your layout shows that there's not much space to waste.

Our litte discussion here has slowly drifted towards a 2-band parametric EQ. Will you make your board available in your shop? It could come in handy.

I will list the latest board revision (including volume boost) in the shop next week.
If you want one (or two) of the first revision (without volume boost) let me know and I will just send it to you.

[kothoma]

If you want one (or two) of the first revision (without volume boost) let me know and I will just send it to you.

PM'd. Thanks a lot!

[RobA]

Not sure if you want all the fun for yourself, so I added a spoiler :-)

Only the  boost/attenuate and the volume pot are outside. Q and Freq adjustment are inside via trimmers.


SPOILER
-----------------------------------
Schematic & Layout:
http://diy.thcustom.com/wp-content/uploads/2014/09/1-band-parameq_V1.0.pdf

One of the prototypes:

The volume control is a good idea. I've noticed while playing with the graphical EQ circuit, that there are definitely EQ settings where you could use some volume boost or cut to keep the overall perceived volume where you want it. I'm not sure about the right way to do this for a Swiss army knife utility board though. In many cases, it'll be handled be the volume control of the main circuit, so it'll need to be optional. I'll think about this. It could be that a simple BJT output gain stage with a preceding volume pot could do the job. If you want to skip the volume section, you leave off all the associated parts to the amplifier and take the output of the volume pot input pad. If you do want the volume, you take it from the output of the final cap. That would add more size to the PCB though.

[RobA]

... Meanwhile I'm having great fun playing with values in your LTspice model!

I've found that it can be a really useful tool for getting ideas worked out even with its limitations. It can be fun to play with too.

[kothoma]

The volume control is a good idea. [...] there are definitely EQ settings where you could use some volume boost or cut to keep the overall perceived volume where you want it. [...] It could be that a simple BJT output gain stage with a preceding volume pot could do the job. [...] That would add more size to the PCB though.

Hm, how about avoiding an additonal gain recovery stage by making the input buffer (optionally) into a small (fixed) gain stage and only tack on the volume pot at the end?

[RobA]

The volume control is a good idea. [...] there are definitely EQ settings where you could use some volume boost or cut to keep the overall perceived volume where you want it. [...] It could be that a simple BJT output gain stage with a preceding volume pot could do the job. [...] That would add more size to the PCB though.

Hm, how about avoiding an additonal gain recovery stage by making the input buffer (optionally) into a small (fixed) gain stage and only tack on the volume pot at the end?

That's a possibility and I'll give it a test too. The concern I have with doing it that way would be that adding the gain before the EQ and the cut after the EQ might cause distortion in the EQ op-amp with some boost settings.

[RobA]

I've done a bit of breadboarding on this now, and it does look like all the modular possibilities will work. The one thing that's still a bit of a question is where to put the gain. The cleanest place circuit wise is to add a bit of gain on the first buffer op-amp. But, you've got to be pretty careful here. Using the TS922, I can push a gain of about 6 times without distortion, that's using some pretty hot P90's. OPA2134's had distortion at a gain of around 3 times, but they tend to not be too happy in single-supply circuits to begin with. I'd say that if you only want a gain of two to four or so, this would be an OK spot to do it. (BTW, without distortion means I can push a really high Q on the PEQ band and peg the high shelf and have the added gain from the first stage with out distortion.)

If you are driving the EQ hard from a preceding stage, it could end up with trouble, but then you probably wouldn't want any gain here to begin with. The cute thing about doing the circuit this way is that if you jumper the feedback resistor and leave off the resistor and cap to ground, then you've got a straight unity gain buffer and everything is cool.

I'm going to think about the circuit just a bit more and then I'll layout a PCB for it and we can give it a try.

[kothoma]

I've done a bit of breadboarding on this now, and it does look like all the modular possibilities will work. The one thing that's still a bit of a question is where to put the gain. The cleanest place circuit wise is to add a bit of gain on the first buffer op-amp. But, you've got to be pretty careful here. Using the TS922, I can push a gain of about 6 times without distortion, that's using some pretty hot P90's. OPA2134's had distortion at a gain of around 3 times, but they tend to not be too happy in single-supply circuits to begin with. I'd say that if you only want a gain of two to four or so, this would be an OK spot to do it. (BTW, without distortion means I can push a really high Q on the PEQ band and peg the high shelf and have the added gain from the first stage with out distortion.)

If you are driving the EQ hard from a preceding stage, it could end up with trouble, but then you probably wouldn't want any gain here to begin with. The cute thing about doing the circuit this way is that if you jumper the feedback resistor and leave off the resistor and cap to ground, then you've got a straight unity gain buffer and everything is cool.

I'm going to think about the circuit just a bit more and then I'll layout a PCB for it and we can give it a try.

That's good news. I'd say the gain is only needed to allow for unity gain not as a boost per se. So it's mostly needed on heavy cut settings. At least it's a nice option to have that can easily be left off as you said. A gain of up to 4 times should suffice. And one could always add a little booster board at the end if needed. I think you're on the right way. I'm really excited to see how much space all this will take.