ive built an lm1875 gainclone and after much tweaking i have to conclude that it just cannot sound 'right' below 20x gain as specified by the chip's datasheet. I imagine the thing will sound even better (read: more stable) at 30x gain or more, but the thing is i dont need all that gain and frankly will feel uneasy knowing that all the noise before the gainclone is being amplified unreasonably when i only ever need 1/3 of its output, not to mention the amp is now set to clip and blow the speaker if the volume pot is ever accidentally turned to max. And my preamp is opamp based so it already needs 3x gain to stay stable.
So does it make sense to put a fixed attenuator at the input of the amp if im not too concerned about messing up the input impedance of the lm1875 since i plan on buffering it anyways?
So does it make sense to put a fixed attenuator at the input of the amp if im not too concerned about messing up the input impedance of the lm1875 since i plan on buffering it anyways?
Absolutely, no problem at all. If the preamp has an output coupling capacitor, keep the attenuator impedance
high enough to set the LF corner below 20Hz (if it is a film capacitor), or 2Hz (if it is an electrolytic capacitor).
If there is no coupling capacitor, just use a 10k (total impedance) attenuator.
Last edited:
Or change the non-unity gain stable op amp to the version that is, and change the feedback
to get unity gain. For example, use the OPA627 instead of the OPA637, etc.
to get unity gain. For example, use the OPA627 instead of the OPA637, etc.
Thank you. And i cant figure this out on my own- does the SNR of all previous stages combined remain constant or not going into the input attenuator of the amp vs the amp without the attenuator? Or does it actually benefit the signal SNR? I cant tell what the negative effects of the attenuator would be besides the added noise of the resistor.
I have messed with buffer preamps before settling for the current one of which sound i like the most (its a low powered whammy headphone amp if you must know).
And reading around it looks like 30x gain for 1875 is recommended, at which point i will definitely need to attenuate
And reading around it looks like 30x gain for 1875 is recommended, at which point i will definitely need to attenuate
Last edited by a moderator:
It's hard to say about the total noise, if you both raise the amplifier gain, and lower the input levels.
Certainly, the noise from all sources would be lower after passing through an amplifier input attenuator,
but then the amplifier with higher gain would amplify the remaining noise more. It's doubtful that
the attenuator would contribute significant noise itself in this case.
The amplifier's own residual noise could be higher if its gain is increased. Should be easy enough to try it, though.
Last edited:
Attenuators can be a source of noise if not low impedance, but a quick look at the LM1875 specs shows it has an input equivalent noise of 3µVrms CCIR, which is probably a lot noisier than a 10k attenuator from my quick calculation.