I had the same suspension, but what I really wanted with the thread, was to hear other people's experiences and alike.
That why I have the four pair of MJ15003/4 in mind, to make an amplifier with enough "power surplus".
I get it. When I simulate amplifiers using spice, I can easily get a good result if my load is a 4 ohm or even a 2 or 1 ohm resistor, but when adding a 100nF capacitor I get ringing and stability problems. I gues the reason is that when it takes time to raise the voltage on my load, the feedback loop will try to encrese the voltage even more and then overshoots and "gets confused".
I have thought of stating the feedback loop on the output drivers instead and thereby sacrifice some DF (as the output transistors then will be outside the feedback loop) but at the same hand get som stability. I'm not shure if I get the same result ig I just add a 0.1 ohm resistor to the output :-D
You mentioned that a sealed box was better than an ported og in my case using passive radiators. First I have to mention that the little SB with the passive and a LM3886 does a fare better job than when I used a Peerless in a sealed box.
Also if I had the skils, I would go for a servo woofer instead. I have once used a feedback loop that used the current output aswell and thereby got a -1 ohm output impedance. I couldn't hear any difference, but that was properbly due to a poorly chosen driver
My plan is to use this trick in mid/tweeter section. A few years back I used non (global) feedback design. They did a good job (just not in the bas area), but over time the DC working point shifted and my DC servo gave up. I know Densen and Holfi uses the same philosopy, so I'm jus not "good enough".
I know. Distortion isn't the issue in the lower regions. I'm actually in doubt that it's what "my ears" are listening for these frequency range. Maybe "my ears" are more into impulse control there and not distortion? Bad impulses are also distortion I suppose.
No worry about the cabling. The amplifier will be built in the back of the speaker with 40cm (16inch) off thick cable 🙂
What I don't get, is why 300 watt will sound better, if I only use 10 watt. I'm not argue against it, I'm just wondering. Is it headroom or is it the fact that usually a 300 watt amplifier have a bigger power supply than a 20 watt amplifier and more output transistors and so on?
The amplifier with feedback loop electrically is plenty fast and flat across the audio Band Width plus some.
The driver is another story. At “low” frequencies the driver diaphragm moves as a piston. As the frequency increases standing waves or ripples begin to form. At higher frequencies parts of the diaphragm move independently. The diaphragm has stored energy that continues to move things even after the signal is removed. When the diaphragm is more orderly feedback helps. The driver is much more at fault than the amplifier.
I think of the driver more as a generator than a microphone.
Sitting in the cage there is an APx555 analyzer and an APx1701 transducer test tool. Looking at driver impedance and phase plots you can see well behaved drivers and others that do not test nearly so well. Stored energy and cone breakup shows up clearly in the test plots. It is all about the driver performance, not the amplifier.
Thanks DT
The driver is another story. At “low” frequencies the driver diaphragm moves as a piston. As the frequency increases standing waves or ripples begin to form. At higher frequencies parts of the diaphragm move independently. The diaphragm has stored energy that continues to move things even after the signal is removed. When the diaphragm is more orderly feedback helps. The driver is much more at fault than the amplifier.
I think of the driver more as a generator than a microphone.
Sitting in the cage there is an APx555 analyzer and an APx1701 transducer test tool. Looking at driver impedance and phase plots you can see well behaved drivers and others that do not test nearly so well. Stored energy and cone breakup shows up clearly in the test plots. It is all about the driver performance, not the amplifier.
Thanks DT
Do I understand you correct, if I translate your post to: "There's no need to change current design with LM3886 into a new amplifier design"
You need enough current to keep the rail voltage firm at full power. larger transformers have better regulation than small VA transformers, less rail voltage sag.
I am more interested in a large enough power transformer to run cool at full power. I want the power transformer cool enough to touch.
Thanks DT
I am more interested in a large enough power transformer to run cool at full power. I want the power transformer cool enough to touch.
Thanks DT
Powering the TDA7294 / TDA7293 chip with bipolar transistors. You can 3886+MJ.
"Особые" схемы усилителей на TDA7294 / TDA7293 - AudioKiller's site
"Особые" схемы усилителей на TDA7294 / TDA7293 - AudioKiller's site
3886: 68W Cont. Avg. Output Power into 4Ω at VCC =
amplifier capable of delivering 68W of continuous ±28V
Output Protection from a Short to Ground or to the supplies, thermal runaway, and instantaneous
to the Supplies via Internal Current Limiting temperature peaks.
amplifier capable of delivering 68W of continuous ±28V
Output Protection from a Short to Ground or to the supplies, thermal runaway, and instantaneous
to the Supplies via Internal Current Limiting temperature peaks.