This thread is a follow-up to this one and that one where I asked for solutions to feed a signal from a guitar pedal device rated 10 dBV (9V p2p) and 10 kOhm output impedance, into an Electrosmith Daisy seed. In the first thread I ask for general considerations regarding impedance, particularly with guitar; in the second thread I asked for clarification regarding a ready-made circuit I found for the Daisy Seed.
Combining what I've learnt from both threads, I decided I wanted to let my device rated 10dBV feed into the circuit without having to turn down the volume knob really low and instead leave the volume knob at about noon. I'd still like to avoid frying my Daisy Seed, so I should ensure the output will not damage the 5V tolerant ADC.
So the solution I've opted for is the following:
Here is what it tries to achieve:
- Inverting Op-amp circuit with a gain of 0.38, which would scale the 9 V p2p down to 3.42 V p2p, as the Daisy Seed ADC expect 3.6 V p2p
- Input impedance of the circuit is 1 MOhm, which is consistent with good practice in the world of guitar pedals: 1 MOhm in, 10 kOhm out
- Power is achieved with a 5 V buck converter fed with a power supply rated 12 V DC, 2.5 A which is also used to power the Daisy Seed
- Op-amp is powered with 5 V because according to bobflux who very kindly and helpfully replied in my second thread, the inverting amplifier setup allows it, although the input signal is 9 V p2p, and the added benefit is that the output cannot exceed 5 V and damage the Daisy Seed
- Bias is also 5 V, the 9 V p2p input should oscillate between 0.5 to 9.5 V
- U12 and U13 deal with high frequency interference (as learned in my second thread)
- C1 and C4 are DC blocking, the values I took from this site about guitar pedals
- C2 and C5 gradually rolls off the high frequencies above the audible range according to same site, I wonder if they are now redundant with U12 and U13
- Although a guitar is mono I'm adding the option for stereo since there are two inputs, hence the same inverting op-amp setup repeated twice
My question is now: will this work ? I think I've applied everything I've learnt, but I am not sure that it is correct. I am especially unsure about the combination of +5 V /0 V power to the OPA1678 and 5 V bias to the +input of the Op-amp: I thought the 9.5 V peak of the incoming signal will clip the Op-amp and distort the signal. I trust bobflux's knowledge, but any deeper insight on why this works would be very appreciated.
Thank you very much for all the help!
For completion, let me summarize all the activity in this thread.
- Justme very correctly pointed out I should bias with Vcc/2 and that I should not use a buck converter to power an op-amp intended for an audio application, bobflux also warned me in an earlier thread. A better approach is an LM7805 linear converter.
- GodJihyo provided a non-inverting setup which should just work, so I validated the answer.
My inverted op-amp setup might just work though, as bobflux was explaining in the comments to his reply on my second thread. I have been mislead and confused a bit by simulations which emulate behaviors of certain op-amps which do not allow operation on the full Vee to Vcc range: some op-amps apparently will start clipping at Vee + 2 and Vcc - 2 and the simulations led me to believe momentarily that bobflux's suggestion wouldn't work since I had 1V left to work with when applying +5V/0V to the op-amp. The OPA1678 recommended by bobflux can output between Vee + 0.8 and Vcc - 0.8, leaving me 3.4V of headroom which is just about what I need, before any clipping happens. I think therefore it would be possible to apply 5V at Vcc and put Vee to ground, while using a voltage divider to divide Vcc by 2 and apply that to the +input of the OPA1678. The result would be indeed an inverted op-amp which scales 9V p2p down to 3.4V p2p and that would saturate at 3.4V p2p.
Here is the diagram:
I would like to thank everyone very warmly for giving their inputs.
