EQ Frequency Reponse - Adjustments needed

Carnhill inductors arrived today and I finished breadboarding this circuit.

As you can see, I need to narrow the Q on several of the mid frequency bands, and work on the amount of cut per band.

This is a bit odd. The fact you have not enough cut implies the series resistance of the inductors is too high. The fact that the Q is not narrow enough implies the inductor values are to small. So you need to increase the inductor values but this will increase the series resistance and give you even less cut.

Can you post the schematic please?

Cheers

Ian

It’s the one you helped me work out a week or so ago.

Here is the schematic:

https://www.dropbox.com/s/svgkb3f45u9msgz/EQ%2C%20Three%20Band%20%28Draft%2003%29.pdf?dl=0

I'm working it out. Moving up to higher inductance and will adjust the series resistors accordingly. Messing around so far is fixing the wide Q. Starting to look nice.

Will report back later.

To get the maximum cut right you will probably have to tweak the cut half of the resistor ladder. In the REDD EQ, the cut inductors are separate from the boost ones so as to maintain the same Q in cut as in boost. The cut inductor values are much smaller and so their dc resistance is lower.

In the Helios 69 EQ, where the same inductor is used for both boost and cut, its series resistance is higher so the bottom resistor in the ladder has to be adjusted to compensate. In my Helios 69 design the resistor value is 470 ohms compared to 1K4 in my REDD design.

Cheers

Ian

Okay, so now I have another question. How can I balance the Q of the boost versus the cut?

Just measured my 1073 EQ, which has equal Q whether boosting or cutting, and I always love the sound of it.

I'll keep searching around, but just in case anyone want to chime in.

I just saw the thread about using two pole switches to balance the LCR circuit. Back to the drawing board.

https://groupdiy.com/index.php?topic=61945.msg800434#msg800434

Okay. I got it worked out now. Still has broader Q for boost and narrower for cut, but it's much more usable now.

New values are:

220Hz = 1H, 0.52uF
360Hz = 1H, 0.2uF
700Hz = .6H, 0.082uF
1.6kHz = .3H, 0.033uF
3.2kHz = .1H, 0.025uF
4.8kHz = .1H, 0.01uF
7.2kHz = .1H, 0.048uF

Measured them in real world with frequency sweeps and everything looks much better. The cuts go down to -10dB in just about all cases (one or two frequencies only make it down to about -9dB, but I'm not gonna lose sleep over it).

Before and after at 220Hz. Much better!

BluegrassDan said:

I just saw the thread about using two pole switches to balance the LCR circuit. Back to the drawing board.

https://groupdiy.com/index.php?topic=61945.msg800434#msg800434

Click to expand...

In the Helios and REDD EQ the pot divider is basically 22K and 5K1. For boost you are connected across the 22K and for cut across the 5K1. The characteristic impedance for cut is thus much smaller than it is for boost which is why the cut curves are much sharper. If you changed the pot divider to,say, 5K1 and 5K1 the characteristic impedance would be the same for cut as for boost but the boost maximum boost would be no more than 6dB. If you want more boost then either the Q will be higher for cut or you need separate (different) LC components for cut.

The Neve achieves the same Q for cut as for boost by virtue of the circuit topology using negative feedback.

Cheers

Ian