5-band EQ. Gyrators & Boss GE-7. Low noise amps choices?

[nonost]

Hi! A friend ask me for a 5-band EQ for bass. I'm using the Boss GE-7 as a start point. Here the schematic:

https://www.hobby-hour.com/electronics/s/ge7-equalizer.php

It uses op-amps configured as gyrators for each EQ band, and a couple input&output op-amps. There's a lot of garbage regarding the switching though...


I've learnt quite a few things from these posts. Mainly from Rob:

https://www.diystompboxes.com/smfforum/index.php?topic=127554.0

https://www.diystompboxes.com/smfforum/index.php?topic=123011.0


I'm using a dual amp as input&output buffer. As for the gyrators, a quad op-amp plus a dual op-amp (which just one side being used). TL072&TL074 come to my mind as first choices. But I think it's worth it to use a better op-amp regarding noise. Which alternatives could I use? I can't easily find quad op-amps with JFET inputs and better noise numbers. Dual op-amps are easier, OPA2134 looks good and reasonable priced for the input&output op-amp. I'm looking for JFET inputs as Rob suggested in the post linked above.

I'm using a bit of gain for the input op.amp, I think that would be a good idea. The output buffer is unity gain with a vol pot at the end.

I'm using 10k W-taper pots and values close to the Boss GE-7. Here a schema of a gyrator:



For the Q points, I'm aiming to 3.5 for low bands and around 3 for the 1.4khz and 6khz.

Any ideas/suggestions are welcome!

[GibsonGM]

I believe NE5532's work well in the GE-7...

[amptramp]

If you are building a unit for bass, you can expect to go an octave lower than the Boss pedal so you may want to add a 50 Hz stage, but as long as the resistor connecting the non-inverting input of the gyrator remains at 100K or lower, you won't get more than a tenth of a volt output offset due to bias current using a 5532.  This is negligible in this design.

But the current noise could be a problem.  The 5532 is quoted as having an input current noise of 1 pA/Hz^(0.5) which is fairly high.  If there was no connection of noise voltage through the capacitors in the gyrator, the noise would be high since connection to a 100K resistor would gives 100 nV/Hz^(0.5) but the gyrator stage noise performance depends on the return of current through the capacitors and the pot to the circuitry around IC4.  By comparison the current noise in a TL072 is 80 fA/Hz(0.5) which translates into 8 nV/Hz(0.5) when connected to 100K but this is also reduced by the circuitry through the capacitors and the pot to the circuitry around IC4.  The current noise sources are not correlated with the voltage noise sources so they add in a square root of the sum of the squares fashion.  The 5532 would be an improvement in noise performance for IC4 and depending on the pot values, slightly better for the gyrator stages.  The 5532 was designed as a mixer board amp for use where the source resistances were on the order of 10K and current noise could be tolerated.

You will get an improvement with the 5532 over the TL072 but not in the same proportion as the ratio of voltage noise.  To go lower in noise, you need a low noise JFET amp.  The OPA 2134 not only has lower voltage noise (8 vs. 18 nV/Hz^(0.5)) then the TL072, it also has a tenth of the current noise, so this would be a good choice.

If you are making this unit on you own board, make the area for the gyrator capacitors large enough that you would not need electrolytic capacitors (and I include tantalum and niobium capacitors since they are electrolytic).  Film capacitors would give you better tolerances and lifespan.

[nonost]

Thanks a lot Ron, I truly appreciate it.

Yes, the idea is to use a band at 60hz as well as some others.

Ok. So OPA2134 for input&output buffers. Would be worth it to use it for the gyrators as well? I only have access to 10k and 20k W taper pots.

What about the LM837? It's a quad op-amp that looks as the 5532

Yep, I'm using only film caps for gyrator caps.

Cheers!

[FSFX]

I was wondering whether anyone had experience using a dedicated 5 channel gyrator EQ chip like the Samsung KA2223, Mitsubishi M5226P or Sanyo LA3600 rather than multiple op amps as used in some pedals.
These are all different manufacturers' equivalents and contain the circuitry for 5 gyrators and an output summing amplifier.
This would seem an easier solution to using five discrete op amps for the gyrators as these devices are still obtainable.

[antonis]

I've built 4 X 5 band and 2 X 10 band (overkill..!!) EQs based on BA3812L with acceptable results..
(taking into account the fact that BJT-based gyrators have higher equivalent output resistance, hence lower gain and lower Q..)



P.S.
To me, identical circuit using op-amps made no audible difference..

[antonis]

A friend ask me for a 5-band EQ for bass. I'm using the Boss GE-7 as a start point.

Why not the Boss GE-7B..??
https://www.hobby-hour.com/electronics/s/ge7b-bass-equalizer.php

[eh la bas ma]

I was wondering whether anyone had experience using a dedicated 5 channel gyrator EQ chip like the Samsung KA2223, Mitsubishi M5226P or Sanyo LA3600 rather than multiple op amps as used in some pedals.
These are all different manufacturers' equivalents and contain the circuitry for 5 gyrators and an output summing amplifier.
This would seem an easier solution to using five discrete op amps for the gyrators as these devices are still obtainable.

I am waiting for parts to start this build :


https://dirtboxlayouts.blogspot.com/2021/10/guyatone-ps-024-distortion-h.html

You can see a direct application for these "5 element graphic equalizer IC".

I am having high hopes about the M5226P i bought online from an electronic shop, two hours away from home (10 euros...). I never heard  these EQ chips before, nor have i seen an other layout with them. I chose Mitsubishi's IC thinking it was a great brand, but I don't really know.

If Guyatone used it, it must be great ?

Edit: You can also consider this build,  Spaceman Effects Equinox clone : https://www.pedalpcb.com/product/pcb527/

[nonost]

Dedicated chips look great! I wasn't aware of them. I'd like to use one of them for the next time.

eh la bas ma, let us know how the thing goes with that chip. It would be great!

Cheers!

[eh la bas ma]

eh la bas ma, let us know how the thing goes with that chip.

Here is the PS-24 build report with a demo. I'd say the EQ is interesting, i like it, but not as much as a GE-7.

https://www.diystompboxes.com/smfforum/index.php?topic=36392.msg1265905#msg1265905

[nonost]

I can't get enough 10w taper pots. I can get a mix of 10k and 20k, do you know if the value is important here?

Cheers!

[MrStab]

Note the gain caused by IC5, R28 and R31, and the attenuator formed by R33 and R36, which provide a handy wee noise win as-is.

I've used LA3600 once for full-spectrum audio, and it was more than permissible IIRC, but pricy, obsolete and limiting beyond frequency selection.