I think we need some more FBI guys...
Scott..want a resistor?
I think I'm ready to semi-bin dive, I'm looking to make some quality differential IA's, so might be lookin for a few IC's...
jn
Quasi-active works the best for me , i don't like what the extra electronics do to the middle , for bass yes , full active here is best. I find multi amping between mid/twt and issue , there's a disconnect thats not apparent with one amp and passive control ...
I have not heard a commercially available active speaker to make me move beyond current stasis.
I think we need some more FBI guys...
Scott..want a resistor?
I think I'm ready to semi-bin dive, I'm looking to make some quality differential IA's, so might be lookin for a few IC's...
jn
Good! maybe now we will get that pre , scotty wont build ...
Last edited:
Good! maybe now we will get that pre , scotty wont build ...
Donna looka me....I ain't doin it...
I'm gonna build a dual channel IA set for my own nefarious reasons. bwaaaaHaaaaHaaaaHaaaa.
Sumptin ta do wit biwiring.
jn
JN,
Interesting measurement, but to translate this into an advise to don't use solid conductor coils goes a bit far.
Your measurement is on a coil with 15 AWG wire. That's a pretty fat coil, so the only use in a crossover would most likely be on the woofer. In that case, the corner frequency where resistance goes up is well above the relevant area. And the effect of the rise in resistance is to allow even less high frequency through than the L part would already do. This is completely benign as far as I can see.
In short, no problem with solid wire inductors when used within reason.
You are absolutely correct. That 5 mH coil was the only one I had the comparison data readily available for.
To do it right, it is important to make the coil and test it in the applicable range. That includes gauge of the wire, aspect ratio, final L value, and range of use.
Using litz will eliminate the vast bulk of the problems, but there is cost involved.
I do note that if my excessively large coil were used in series with a woofer that was built using copper shorting rings, then any non linearities out of the woofer band caused by the inductor that the shorting ring tries to "short" would be impressed across the mid/high circuit.
All these questions are basically why we engineer the solution..
jn
I think we need some more FBI guys...
Scott..want a resistor?
I think I'm ready to semi-bin dive, I'm looking to make some quality differential IA's, so might be lookin for a few IC's...
jn
I think I can set you up. Didn't we go over that absolute value of dI/dt stuff a while back? I would like to see the seconds isolated in an ordinary speaker context, like say a crossover coil with copper plane there and then not there while playing at normal levels. MRI's take it up a notch.
Absolutely did.
I'm co-winding a coilset using 11 awg magnet wire and 30 gauge kynar, specifically to isolate out Rs times I. But I'll have to think about how a conductive plane eddy current will present. Since it occurs via the flux lines, my test design may cancel it out of the measurement.
hmm, gonna have ta kill some brain cells on that.
jn
Edge affect..... if you design pcb for very fast rise time data signals and/or high freqs, the edge affect on Z and delay of those lines have to be taken into account. With the best practices in SW pcb design this is included for Z and timing control. You might look at those pcb sources for more info on the edge affect and how to over come it.
Secrets of extreamist high-end audio -- On my pcb, where there are high currents at HF, I leave the edge of the trace/plane unmasked so that the trace is thicker at the edges after solder flow/reflow. You can also selectively plate the trace edges for thicker/lower HF resistance. [I have done several tests along these lines using HP Z analyzer and found other ways to normalize the Z and thus maintan constant low Z and low group delay over wider range of freqs than just stock pcb wire/paths..... other pcb and other situations]
Thx-RNMarsh
Secrets of extreamist high-end audio -- On my pcb, where there are high currents at HF, I leave the edge of the trace/plane unmasked so that the trace is thicker at the edges after solder flow/reflow. You can also selectively plate the trace edges for thicker/lower HF resistance. [I have done several tests along these lines using HP Z analyzer and found other ways to normalize the Z and thus maintan constant low Z and low group delay over wider range of freqs than just stock pcb wire/paths..... other pcb and other situations]
Thx-RNMarsh
Last edited:
The sense lead inductance shouldn't come into play. Both sense leads have to be clear of the power leads, twisted to prevent coupling. Also, the sense lead terminations within the power amp should be high impedance loaded, 100 ohms to 150 ohms, that way it's settling time is roughly one transit time.
jn
Test.
As long as you don't try for a bandwidth consistent with the length of the cable, I don't see any issue.
jn
Conventional engineers, working for a wage, will never deliver top notch audio. All they worry about is meeting the 'requirements', a few measurements on the 'right' instruments - a good set of numbers - and they're happy ...
This is a game that still requires genuine passion to get anywhere - a brilliant, but bored engineer will serve up Yet Another ho-hum product ...
Thank you Pavel for the screenshots, they are spot-on.
I agree with the rest of your post.
Does this termination advise apply to the feedback network too (assuming the amp has GNF)?
George
I was only speaking about the long cable to the speaker.
Internal to the amp, everybody just runs wires with no worry.
jn
One could argue that only conventional engineers have the skills to meet real engineering requirements - such as amplifying a signal without messing with it too much.fas42 said:
Of course, if there are non-engineering requirements set by Marketing or Fashion and these take precedence with the intended customer base then it is best to avoid engineers as they are not always very good at deliberately ignoring facts. Better to recruit someone who doesn't know the facts; ignorance can be bliss.
Yes, I agree, that's precisely my point - in audio, people can't specify the right requirements - hence, engineers don't have a clear, well-defined target ...
- Status
- Not open for further replies.