John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Jump to Latest
#50,081
Cal,
I agree but what happened to my last post where I asked a real question?

Reposting that.

For John or anyone else versed in Feedback theory. What would happen if we took a third wire back to the speaker terminal and used this as the end point of the feedback loop for a power amplifier, the end point of the gnf point that was used to feed the input node? This would include all wire parasitic value and would get you to the point where you would see the back emf coming from the speaker terminal. Would this have any advantage over the current method of leaving the speaker node out of the feedback loop?
 
#50,083
What about EMC pick up on the wire? I believe that that could be a possible problem, unless the amp was very near the drivers, say for an active system.
Conducted susceptibility with the wire acting as an antenna could be a problem, and each application would require a different length wire tailored to the users set up. For low level designs feedback loops are usually pretty sensitive, and loop area is best reduced to a minimum, don't know for power amps though, how sensitive would they be?
 
#50,085
Kindhornman said:
Cal,
I agree but what happened to my last post where I asked a real question?

Reposting that.

For John or anyone else versed in Feedback theory. What would happen if we took a third wire back to the speaker terminal and used this as the end point of the feedback loop for a power amplifier, the end point of the gnf point that was used to feed the input node? This would include all wire parasitic value and would get you to the point where you would see the back emf coming from the speaker terminal. Would this have any advantage over the current method of leaving the speaker node out of the feedback loop?
Click to expand...

Furget that, What about amps with user adjustable. Feedback and Bias ..... :drink:
 
#50,087
a.wayne,
I would thin that would be a dangerous proposition with the common user, perhaps a professional who understood what they were adjusting you could get away with that, but just giving an end user another knob to adjust would probably cause more problems than anything else. Probably see some fried amplifiers I would imagine unless you could limit the range of adjustability.
 
#50,090
I would think so unless you shielded everything, the most common problem we are seeing is wires coming into boxes on a wide variety of designs, so much so that in some cases a quarter of the board is ending up with EMC and transient protection.
I think it would be a possible problem.
 
#50,092
Kindhornman said:
C What would happen if we took a third wire back to the speaker terminal and used this as the end point of the feedback loop for a power amplifier, the end point of the gnf point that was used to feed the input node?
Click to expand...

Just a third wire would be a problem. You will be referencing that third wire to ground, and there will be a loop between that wire and ground.

To properly get fb from the speaker, you need a twisted pair. And, you have to use a true differential input from that pair, as neither will be at ground.

I recall an amp design from back in '74 that did exactly that. The only concern I can think of is the problem of a disconnect of one of the fb wires..to that end, connect a 100 ohm resistor from the output terminal to the fb wire associated with it so that if the wire disconnects, the fb would still be working through the resistor.

If there's any length involved, this wire could be terminated at the amp at it's characteristic impedance, so would never resonate at rf, and the added bonus is that it would be a zobel type termination for the main leads to the speaker.

jn
 
#50,095
Kenwood sold a commercial product that included feedback from the speaker terminals back in the 1970's or 80's. IIRC the package included a dedicated two pair of wires per channel of "fixed" length.

Reminds me of Paul Klipsch's recommendation about speaker wire lengths: if you need to cut a speaker wire and don't know the exact length, always cut it a little short. That way, you can always add on a piece. But if you had cut it too long, what could you do?

Try the veal. Remember to tip your waitresses. We'll be here through the weekend.
Chris
 
#50,096
Kindhornman said:
Cal,
I agree but what happened to my last post where I asked a real question?

Reposting that.

For John or anyone else versed in Feedback theory. What would happen if we took a third wire back to the speaker terminal and used this as the end point of the feedback loop for a power amplifier, the end point of the gnf point that was used to feed the input node? This would include all wire parasitic value and would get you to the point where you would see the back emf coming from the speaker terminal. Would this have any advantage over the current method of leaving the speaker node out of the feedback loop?
Click to expand...

It has been done using a feedback coax with BNC connections. Didn't improve things for am obvious reason that most folks miss. The loudspeaker was voiced using a regular amplifier.


Now for the peanut gallery a couple of simple questions:

What is the maximum output voltage of a compliant CD player?

What is the minimum input impedance of an IEC compliant consumer amplifier/preamplifier?

What is the typical headroom when playing a listening test or high quality CD?

Allowing for the above what is the typical current in an interconnect cable?
 
#50,098
Pavel,
Sometimes it is good to understand the theoretical in-order to understand a practical application and if the practical is even possible. If theory says something is not possible why look for a practical application, that would seem to be a complete waist of time.
 
Status
Not open for further replies.
Top Bottom