Quote:
Originally Posted by Alex Pogossov
"flipper" wrote in message
-------------------
Even if an amp is stable without the shelving network, shelving should
be used especially where there is some peaking in the region where "we
don't intend to amplify anything that low anyway" as you say. The
shelving improves overload charactecter and saturation effect
behaviour.
Yes, I'm sure the OP will appreciate improved overload characteristics
at frequencies that will never be presented to the amplifier.
Do you want rumble (from say a pick-up playing a warped record) driving your
amp to saturation, if your amp is close to LF oscillation?
|
One solution is the use of a bandwidth limiting network on the front end of an amp. I too at that time long ago (40 years!!) had similar concerns. First of all in those days an important program source was the turntable & still is with many listeners.
These were not always perfect so that rumble was a concern.
Another concern had to do with the possibility of driving the loudspeaker at frequencies below it’s normal operating range, perhaps by the above mentioned turntable rumble. These two problems either separately or together are a good recipe for Intermodulation Distortion both in the speaker & the amp.
I was also troubled by possible damage to the output tubes or the speakers caused by switching transients. As a result, I devised a passive HP filter for the amps front end which is two stage rather than one.
To do this you need to arrange the impedance of the second section to be in the range five to ten times that of the first. That’s how I arranged the gain control in the first section & something of higher impedance in the second section. I did not believe a low pass filter made sense at this point.
Here is the response of the high pass network
High Pass Section- 3 db down @ 20 hz
44 db down @ One hz
170 degree phase shift 12 db / octave