Input range on FV-1 chip

[Dodecahedron]

Hello gentlemen! I'm a bit new here so forgive me if I break any conventions  Anyways, I have a question about the FV-1 reverb (and other effects) IC. I've been looking on other people's designs and noticed that many of them feed the amplified guitar signal into the input pins of the IC. Example: https://www.schematics.com/project/fv-1-audio-reverb-effect-51124/

You will notice that, if my math is correct, the input opamp has a gain of 10. Let's say the pickup output is 500mVpp, although on my pickups sometimes it's up to 1V, and I don't know much about other pickups. This will leave us with signal of around 5 Vpp with potential to go higher. This goes straight into the ADC input which the datasheet (http://www.spinsemi.com/Products/datasheets/spn1001/FV-1.pdf) specifies as max Avdd+0.5, with Avdd(max) = 3.5, so 4 V maximum.

I think of using a rail to rail opamp that is powered with 3.3 v so I don't accidentally give the chip too much voltage, but it seems that some designs get away with it? I was interested if anyone has experience with such a design and how tolerant the chip actually is to overvoltage. Also, if given more than Avdd, wouldn't the signal clip?

[Digital Larry]

Probably a question best asked at the Spin forum where you can get an answer from one of the chip's designers.

I'd certainly expect the signal to clip if it tries to go outside the GND to VDD limits.  The chip does offer a clipping LED to help you set a gain trimpot, if you'd choose to put one in your design.  Other people have put clipping networks (e.g. back to back LEDs) in front of the FV-1.  These might have a smoother clipping characteristic that would make clipping less obnoxious.

[Dodecahedron]

Hmm true about the spin forum. Also thanks for the clipping network suggestion, will look into that!

[MetalGuy]

My experience is the FV-1 will start clipping at around +4dB. As Larry mentioned it's a good idea to include a clipping LED in the design. This way the clipping level can be easily measured.

[ElectricDruid]

My experience is the FV-1 will start clipping at around +4dB. As Larry mentioned it's a good idea to include a clipping LED in the design. This way the clipping level can be easily measured.

+4dB relative to what?

[MetalGuy]

+4dBu = 1.23 volts AC

[ElectricDruid]

Ok, thanks. So why would something operating on 3V clip at 1.23Vrms or 1.736Vpp? I'd expect the ADC to be able to get quite a bit closer to the rails than that. I learned something. Thanks.

Sorry for being a stickler about units, but on its own +4dB just means "x1.6 louder". dBu or dBV have set reference levels, so then you've got a guideline.

I still don't like them much. Why refer to something relative to something else? Why not just call it what it is? If the FV-1 clips at 1.73Vpp, we can say that. For me, referring it back to an (now) arbitrary 0.7746Vrms level and then saying how much gain you have to apply to that to get clipping is just overcomplicated. Others may have different views, obviously, and that's fine. It's not a religion, and my god of volts peak-to-peak isn't going to call me to war with your tribe because your god is decibels or RMS or something.... and thunk fark for that!

[Digital Larry]

http://www.spinsemi.com/Products/datasheets/spn1001/FV-1.pdf

Note that there is an "Absolute Maximum Rating" for analog input voltage, which I believe means: "If you have a low impedance source and go outside this range, you may destroy the chip".  That is the -0.5 to Vdd + 0.5 range and you will be clipping the heck out of your audio before that.

Under "electrical characteristics" you will see "Maximum Input Voltage" shown as between 2.6 to 3.0 Vp-p.

I actually don't know without running down and turning on my scope and all that.  I also don't represent Spin, so I can't tell whether reports of lower clipping levels are typical.  Some people apparently wind up with really old chips being sold by questionable distributors.  This is another one of those things I'd ask about over at the Spin forum since one of the chip's designers will answer your question, which is more authority than just about anyone else.