I am making a guitar interface for the Raspberry Pi. I have got a working setup using a codec evaluation board and a guitar to line level preamp on a breadboard but i need some advice on selecting a good op-amp (SMD).
I am using the virtual ground circuit from this thread (first Picture): http://www.diyaudio.com/forums/inst...uitar-level-line-level-hi-quality-preamp.html
The difference is that i have only got 3.3V and 5V supply from the Raspberry Pi. Currently i'm using a TL072 and it seems to work. However when i read the specs i see that with +/- 2.5V i am way below the supply voltage ratings.
In the second reply member voltwide states that the TL071 (and i assume the whole TL07x range?) is a much better choice as an input stage with a hi-impedance source than a NE5532. Why is that? Because of the JFET input?
Any help would be highly appreciated.
I am using the virtual ground circuit from this thread (first Picture): http://www.diyaudio.com/forums/inst...uitar-level-line-level-hi-quality-preamp.html
The difference is that i have only got 3.3V and 5V supply from the Raspberry Pi. Currently i'm using a TL072 and it seems to work. However when i read the specs i see that with +/- 2.5V i am way below the supply voltage ratings.
In the second reply member voltwide states that the TL071 (and i assume the whole TL07x range?) is a much better choice as an input stage with a hi-impedance source than a NE5532. Why is that? Because of the JFET input?
Any help would be highly appreciated.
OPA1641 is a recent low noise fet input, R-R output op amp that runs on 5V, although the input cm range is off-center
R-R in/out is more common with CMOS op amps whch can be a bit noisier
R-R in/out is more common with CMOS op amps whch can be a bit noisier
Last edited:
found my decade old plot of CMOS op amp input noise - http://www.diyaudio.com/forums/solid-state/68793-new-interesting-cmos-op-amps-2.html#post843381
Would LME49721 be a good choice for this circuit: LME49721 High Performance, High Fidelity Rail-to-Rail Input/Output Audio Operational Amplifier | TI.com
Especially regarding the input caracteristics. The gain and output caracteristics are of lesser importance since it will operate mainly as a buffer and attenuator before the ADC.
Especially regarding the input caracteristics. The gain and output caracteristics are of lesser importance since it will operate mainly as a buffer and attenuator before the ADC.
Or perhaps the OPA130/OPA2130?
OPA130 Low Power, Precision FET-Input Operational Amplifiers | TI.com
OPA130 Low Power, Precision FET-Input Operational Amplifiers | TI.com
LME49721 has the widest input common mode range with 5V supply, better for high output pickups
the OPA130 would only have 1 Vpp input common mode range with the same supply
but the LME is fast at 20 MHz and would require some layout care to prevent oscillation - small ceramic supply bypass C, output series 100 Ohm
the supply current is higher than some other options, but not as much of a consideration if powered from a 5V supply instead of a battery
the OPA130 would only have 1 Vpp input common mode range with the same supply
but the LME is fast at 20 MHz and would require some layout care to prevent oscillation - small ceramic supply bypass C, output series 100 Ohm
the supply current is higher than some other options, but not as much of a consideration if powered from a 5V supply instead of a battery
Last edited:
I would use the LME49721. The only thing you need to be mindful of is the grounding scheme reguardless of what op amp you use, because its easy to create a internal ground loop when both digital and analog share the same ground.
Thank you everyone!
I ended up using a TLV2379 and it seems to work pretty well. Regarding the ground scheme i have got three separate areas on the ground plane: analog, digital and one special area under a noisy 5V track at one edge. They meet only at the star point.
I ended up using a TLV2379 and it seems to work pretty well. Regarding the ground scheme i have got three separate areas on the ground plane: analog, digital and one special area under a noisy 5V track at one edge. They meet only at the star point.
Well, that chip sure does draw extremely LOW power supply current (4uA), but it is about 25db NOISIER than others, and about 660 times SLOWER. Hmmm.....big compromises, methinks.
There are many sources of noise that are far greater than the op-amp in this case, but the low bandwith concerns me a bit. I didn't notice that... :-/
I have got only 5V Power supply and the signal level from my guitar pickups can reach 3.5 to 4V p-p at the transients so i need something with input rail-to-rail to both V- and V+.
> using a TLV2379
90KHz gain bandwidth product.
> circuit from this thread (first Picture)
Gain of 3. Response will be down -3dB at 30KHz, -1dB at 15KHz. "Worse", distortion may be significant above ~~1KHz due to feeble feedback.
> a guitar interface for the Raspberry Pi
Ah, then on response and THD alone, no real problem. The high-cut is useful for keeping supersonics from aliasing down into the audio.
The hiss does look frightening. Of course depending why you want signal in the computer.
90KHz gain bandwidth product.
> circuit from this thread (first Picture)
Gain of 3. Response will be down -3dB at 30KHz, -1dB at 15KHz. "Worse", distortion may be significant above ~~1KHz due to feeble feedback.
> a guitar interface for the Raspberry Pi
Ah, then on response and THD alone, no real problem. The high-cut is useful for keeping supersonics from aliasing down into the audio.
The hiss does look frightening. Of course depending why you want signal in the computer.
- Status
- Not open for further replies.