After much thinking, reading, and wanting to make a good amp that has less hiss (noise) and low distortion, with low DC offset, and less components, I have concluded to make my next power amps fed from the inverting input instead of the noninverting input.
1. The input and feedback capacitors - Mixed opinions about capacitors and distortion and DC offset in the input stages here at DIYAudio, so to eliminate needing two caps (typically 2-10uf input, 100-220uf on the feedback) and just have ONE large value cap (220-1000uf) driven by low impedance preamp or buffer to the inverting input to keep good bass response and no DC offset. The noninverting input on the power amp being grounded permanently so no high impedance input causes noise pickup from surroundings.
2. Less extra resistors and compensation (filtering) needed on input of power amp. Non-inverting input being grounded, so no need for anything there, and being driven by the inverting input instead means low input impedance of power amp, so less frequency compensation is needed to avoid picking up noise, and easier to have flexible range of filter caps instead of just the typical 100-220pf, just in case you want to use larger or smaller ones, and I believe trebles may sound better because low impedance input, and were not filtering away our treble response so much to keep it stable. (better full range sound)
3. Subwoofer-Only use, you can avoid hum, again because low impedance input, but also design frequency compensation into the amp's feedback network itself to reduce voice and treble frequencies, so the amp is filtered down and will never pick up RF, and makes a simple, oscillation-free DIY amp.
4. BTL use - both amps are the same gain resistors and same gain (R1/R2 in the feedback network) so BTL is easy by using same value resistors.
Have you folks done this before, and whether you have or not, I wanted to bring up the discussion, to see some points made by other DIYAUDIO members.
This may no be standard practice, but I like to think outside the box and try new things, so even if other typical designs have been used for years, I still want to always see if there's a better or different way.
1. The input and feedback capacitors - Mixed opinions about capacitors and distortion and DC offset in the input stages here at DIYAudio, so to eliminate needing two caps (typically 2-10uf input, 100-220uf on the feedback) and just have ONE large value cap (220-1000uf) driven by low impedance preamp or buffer to the inverting input to keep good bass response and no DC offset. The noninverting input on the power amp being grounded permanently so no high impedance input causes noise pickup from surroundings.
2. Less extra resistors and compensation (filtering) needed on input of power amp. Non-inverting input being grounded, so no need for anything there, and being driven by the inverting input instead means low input impedance of power amp, so less frequency compensation is needed to avoid picking up noise, and easier to have flexible range of filter caps instead of just the typical 100-220pf, just in case you want to use larger or smaller ones, and I believe trebles may sound better because low impedance input, and were not filtering away our treble response so much to keep it stable. (better full range sound)
3. Subwoofer-Only use, you can avoid hum, again because low impedance input, but also design frequency compensation into the amp's feedback network itself to reduce voice and treble frequencies, so the amp is filtered down and will never pick up RF, and makes a simple, oscillation-free DIY amp.
4. BTL use - both amps are the same gain resistors and same gain (R1/R2 in the feedback network) so BTL is easy by using same value resistors.
Have you folks done this before, and whether you have or not, I wanted to bring up the discussion, to see some points made by other DIYAUDIO members.
This may no be standard practice, but I like to think outside the box and try new things, so even if other typical designs have been used for years, I still want to always see if there's a better or different way.
Last edited:
if you use inverting with only one voltage gain stage, the signal will be 180 degrees out of phase.
Hi,
Look up gainclones, and specifically use of a high quality dual op-amp
to allow no electrolytics in the signal path / feedback. Inverting mode
eliminates common mode distortion, not sure that its always an issue.
Things might go a bit mental if you turn the source off though ......
rgds, sreten.
Bob Cordell :
Look up gainclones, and specifically use of a high quality dual op-amp
to allow no electrolytics in the signal path / feedback. Inverting mode
eliminates common mode distortion, not sure that its always an issue.
Things might go a bit mental if you turn the source off though ......
rgds, sreten.
Bob Cordell :
Attachments
Hi,
Do not forget that if you do not plug its input to a low impedance source, an inverting power stage will work with a null voltage gain at the output and a closed loop gain equal to the full open loop gain. This may lead to bad oscillations and even a failure of the circuit due to inadequate frequency compensation. Using a buffer (like Bob Cordell's circuit) avoids this.
Do not forget that if you do not plug its input to a low impedance source, an inverting power stage will work with a null voltage gain at the output and a closed loop gain equal to the full open loop gain. This may lead to bad oscillations and even a failure of the circuit due to inadequate frequency compensation. Using a buffer (like Bob Cordell's circuit) avoids this.
Forr's suggestion (Cordell's buffer) in effect converts the amplifier to balanced input. Then choose which input to ground and which becomes input.
Cool, thanks for your responses. I'll have my input signal inverted (obviously) to have correct phase.
Since the source impedance has to stay low for stability, ill use a 1k ohm on input to ground just in case it gets unplugged during testing.
The preamp stage will be capable of driving very low impedance load, under 100 ohms.
The volume control will be part of the source (car stereo head unit) but ill have trimmer pots in the preamp to adjust final gain, no volume pot on output amplifier, because it is driven directly by preamp stage through the inverting input.
Since the source impedance has to stay low for stability, ill use a 1k ohm on input to ground just in case it gets unplugged during testing.
The preamp stage will be capable of driving very low impedance load, under 100 ohms.
The volume control will be part of the source (car stereo head unit) but ill have trimmer pots in the preamp to adjust final gain, no volume pot on output amplifier, because it is driven directly by preamp stage through the inverting input.
- Status
- Not open for further replies.