NE571 LF response

[Mick Bailey]

I just built an NE571-based noise gate and it's worked out well, other than a couple of issues. The first is that there's a slight loss of low-end with the pedal engaged and I can't see where the problem lies. The coupling caps should be fine for guitar, giving a maximum corner frequency of 11.2 Hz. Does anyone know if there's a frequency-dependent aspect to the NE571 configuration? I can't see anything that stands out from the spec sheet.

There's also slight hiss from the output with high-volume use, or feeding into a high-gain amp. Swapping to an NE570 improved matters quite a bit, though I think the residual noise may be down to the internal 741-type opamp. Using a low-noise external opamp may be the best answer and this seems to be the preferred way to connect the device for more critical applications. I already switched to an NE5532 for the preamp. Having said there's hiss, the thing that makes it noticeable is the general absence of other noise when engaged and it's at a far lower level than the noise with the gate switched off.

The nice thing about this pedal is there's no obvious gating and when set up just right it completely eliminates all single-coil noise from fluorescent lighting or proximity to a laptop with my Tele without affecting the note decay. At extreme settings it will give a slow attack, but this is easily avoided. I'm quite pleased with this - cheap to build, effective and unobtrusive. I dispensed with the bypass circuitry and switching and built it as a regular true-bypass pedal. Just needs a tiny bit more work to make it excellent.

[nickb]

I don't see why the LF should be dropping off. All the poles seem the be low enough. Have you checked the component values? Failing that scope it out to see where it's falling off.

On the noise it seems slightly sub-optimal to have the gain stage after the first stage.

[Mick Bailey]

Component values are good. I'll check it out with the scope later today. I checked last night that noise is still there with the NE5532 removed and the input to the 570 grounded, so it's coming from the IC. One aspect of the circuit is that because it kills all the other noise the focus is on the noise it makes itself. A bit like living on a busy main road and then being annoyed by a ticking clock the day the road gets closed.

I think it could be better, though.

[Mick Bailey]

Nothing wrong with the LF going into pin 6 of the 570, but on the output pin 10 the signal starts to reduce from just over 100hz. So it's something to do either with the configuration or the IC's internal response. Still scratching my head.......

[strato56]

Does increasing C8 help ?

[Mick Bailey]

I haven't experimented with cap sizes. Easy enough to try though. Is there a specific reason why that particular cap would need to be increased?

[strato56]

Philips uses 2.2µ in their data sheets. Try measuring at pin 7 & 14 also.

[Mick Bailey]

Thanks. I found that increasing C9 (1uf) to 10uf completely fixed the LF response in this particular circuit. If I was building this again I would also increase C5,C6,C7,C8 to 2.2uf as per the Phillips data sheet.

The remaining noise issue is down to the internal reference voltage noise coupled with the internal opamp noise. The datasheet shows that with an external opamp the noise can be filtered out. This along with a quieter opamp would probably be the next step. Both opamps can be externalized to improve performance. I have this incarnation already built up on vero so no room on this board. It's a nice circuit and would justify another build onto PCB.

With a regular medium-gain, medium-volume amp the noise isn't a problem, though. Again, its only because everything is really quiet I start listening to what's left.