LM1875 - Safety about no input connection, with input resistor

[xeroname]

Hello everyone.

Is it okay to leave LM1875's input without connect any signal source, but with RC network like below schematic?

I've heard leaving op-amps input to floating can cause oscillation, abnormal temperature rise in some cases.

But what if there are RC filter acts like pull down resistor like above schematic?
Would it be okay leaving AMPIN_L to no connection because the opamp's input is referencing ground?

Please let me know your opinion.
Thanks in advance!

 

[MarcelvdG]

This should be fine, also with open input: R4 and R5 ensure there is a DC path and C6 keeps the impedance low at high frequencies. I would place R6, just to keep the input DC voltage with no source predictable.

 

[xeroname]

Hello MarcelvdG.
Thanks for your quick and detailed advice!

Actually I'm now struggling with amplifier noise at buffer stage, no doubt your advice will be helpful to me!
Hopefully would be fixed in several days 🙂

 

[tomchr]

+1 for what Marcel said. I'm also not a fan of leaving capacitor ends floating. The 1 MΩ won't hurt anything and it'll keep the input cap from collecting charge.

Tom

 

[xeroname]

Hello Tom.
Thank you for your more specified description!

Could you also kindly explain me whether below diagram should works?

I'm using NE5532 as a unity gain buffer to isolate exeternal signal and keep output impedance of buffer to low.

The key point above diagram is this : There's LPF and HPF networks at LM1875's input but there's only 22kOhm resistor at NE5532's input, without any DC blocking capacitor.
Would this be safe at any source of audio signal, including situation of NC coaxial?

 

[tomchr]

Most amps with a JFET or BJT input will need somewhere for the input bias current to go, so they need a DC path to ground. Your figure shows such a path for both the LM1875 and the NE5532 so I'd say it's good ... at least as far as floating inputs go.

Tom

 

[rayma]

Why do you insist on having an open input? That is never a good idea. Neither is hot swapping.

 

[xeroname]

That's because I want to make durable amplifier, could be treated like any other "commercial " home appliances.
So I thought considering most of unwanted situation is important and would like to reinforce my design through other's advice.

Most amps with a JFET or BJT input will need somewhere for the input bias current to go, so they need a DC path to ground. Your figure shows such a path for both the LM1875 and the NE5532 so I'd say it's good ... at least as far as floating inputs go.

Tom

Also huge thanks to the Tom, that explanation really helps me a lot!
Wish you have a nice day 😀