If it were in 2015, those would be the power supply caps. Yes, some esoteric amps do use 100,000 uF caps in the power supply. Or more. Way past the point of diminishing returns. Personally, I’ve never seen anything over 10,000 uF as an *output* cap, but I wouldn’t have even been paying attention to $10,000 amps back in the capacitor coupled days, so I wouldn’t know what was on the market. I have several with 6800’s, and one with 10,000 uF (home brew, 150 wpc fail-safe lab amp).
As far as power supplies for amps used in the home, 10 to 20K uF (per rail) is plenty.
Those esoteric amps are rediculous things anyway, designed to impress with stats and appearance for the nutcase crowd.
"Oh look at this one!... it runs for 6 months after you pull the plug!" (it must be made by aliens)
Gimme a break.
I'll stick to living in the Real World.
Those esoteric amps are rediculous things anyway, designed to impress with stats and appearance for the nutcase crowd.
"Oh look at this one!... it runs for 6 months after you pull the plug!" (it must be made by aliens)
Gimme a break.
I'll stick to living in the Real World.
For a variety of reasons, 20,000 uF per rail works out to be be right number for two channel amplifiers driving 4 ohm loads. The required capacitor scales with load impedance, just like the output capacitor (if used) does. The only time a bigger capacitor is actually needed is when an amplifier truly needs to be able to operate down to DC. Funny thing is that many amps that *were* designed for that (early Crowns, for example) skimped a bit on the caps and still worked well. A ton and a half of feedback, and input stages designed to limit before the outputs saturate helped cover the resulting sins. More modern PA amps with common emitter output stages and undersized power supply caps sound like absolute garbage when they run hard with low bass. You hear the IM distortion with the power supply ripple. Bump them up to 10kuF per channel and they would sound a lot better. But go too far and you overheat the transformers and blow rectifiers, in addition to pricing yourself out of the market.
During a power outage, I'd MUCH rather be concerned with more important things in life than sitting there enjoying music.
Besides, a battery-operated radio might be the thing to listen to while I move spoilable foods to the freezer for safety.
There's an idea!........ a UPS for the refrigerator!
Why hasn't that lousy, always being sued company Samsung devised such a device?
And as I've said on occasion, who listens to DC?
And why make an amp designed for audio have that ability?
I'm glad you said that. I was beginning to suspect our young friend had mistaken big power supply caps for output caps. That is likely the case.
I read somewhere that the ESR of a big capacitor reduced that damping factor, making the bass sound different. I have no knowledge of whether this is true of capacitor coupled amps, but I have tried a 1ohm resistor ( bypassed by a 1uf cap ) in series with speakers, and the effect was not unpleasant, especially with aperiodic loaded speakers. One reason for having more than 10,000 uf caps on the power supply is if the voltage is low, I've got 4.5volts, and super capacitors are cheep.
A UPS for a refrigerator is called a Honda gasoline generator. We had plenty of practice with that in TX this past winter. A minute or so interruption in power while moving the plug from the wall socket to the generator cord won’t cause the ice cream to melt. Powers the air handler to the gas furnace just fine, too.
Large caps make sense on low voltage power supplies, if current is high and several volts of ripple are ill-afforded. But with an audio amp driving a 4 ohm load on 12 volts, you’re not going to be drawing much current, either, so the ripple should be small. Something with a big DC-DC converter in it might draw a bit more.
Large caps make sense on low voltage power supplies, if current is high and several volts of ripple are ill-afforded. But with an audio amp driving a 4 ohm load on 12 volts, you’re not going to be drawing much current, either, so the ripple should be small. Something with a big DC-DC converter in it might draw a bit more.
A capacitor coupled amp usually implies single rail design and yet I don't think anyone has mentioned the big advantage a single rail design can have over a split rail... and its not that it is speaker friendly or that it is more cost effective or any of those things.
It removes the possibility of negative rail injection of hum/noise/ripple and modulation of the supply by the music.
It removes the possibility of negative rail injection of hum/noise/ripple and modulation of the supply by the music.
I built a single-supply class AB amp with 2-10,000uF's in the 52v supply, and 4700uF speaker coupling caps (stereo).
Driving a pair of big AR LST's at my neighbors house, it performed beautifully, shook the walls and remained clear and detailed.
No need for making some monster-power amp for home use.
In fact, we didn't hear any difference when using his 100/100w McIntosh MC2100 with those speakers.
Driving a pair of big AR LST's at my neighbors house, it performed beautifully, shook the walls and remained clear and detailed.
No need for making some monster-power amp for home use.
In fact, we didn't hear any difference when using his 100/100w McIntosh MC2100 with those speakers.
Have a look on this one. Coupling capacitor included in the feedback loop.
Incredibly low distorsion. Very low output impedance. Protection of the loudspeaker against DC current. Few components. And low cost...
Audio Amplifier Design
THE MJR7-Mk5 LATERAL MOSFET CLASS-AB POWER AMPLIFIER
Incredibly low distorsion. Very low output impedance. Protection of the loudspeaker against DC current. Few components. And low cost...
Audio Amplifier Design
THE MJR7-Mk5 LATERAL MOSFET CLASS-AB POWER AMPLIFIER
The advantage is having speakers that actually work tomorrow after you make a mistake. Like tripping over the speaker wire on a dark stage. Lots of used speaker enclosures with new cheapo drivers at the musician's resale shop "FORCE". I personally had the ST120 amp burn up in a ball of fire under the lecturn at the front of the church, *****y heat sinks and a 3 1/2 hour rehearsal. No damage to speaker due to capacitor. Change amp to a ST70, carry on, Christmas cantata occured on time.
Protection available without a microprocessor or a dozen discrete parts controlling a $40 relay tested to break full rail capacitor energy dump through the speaker. $3 part, a speaker capacitor.
Note my tag line, I'm mostly using MMA-875T and MMA-8150 amps with a speaker transformer. $60 amps, $600 speakers. No damage to speakers it took me 9 months to find @ $400 and a 410 mile road trip to collect them.
Last edited:
The popular ACA here on diyAudio has the coupling cap within the feedback loop. If you plot the response with the 3300uF as recommended or reduce it by a factor of ten to 330uF the result is the same... at least outwardly.
In practice the signal level before the cap at low frequencies is very much larger so as to take into account the high reactance of the lower value cap.
This is using 3300uF and 330uF into 8 ohm. The top trace is actually the two different values overlaid on each other. Essentially identical in response.
In practice the signal level before the cap at low frequencies is very much larger so as to take into account the high reactance of the lower value cap.
This is using 3300uF and 330uF into 8 ohm. The top trace is actually the two different values overlaid on each other. Essentially identical in response.
Attachments
Plus, as an added measure, I usually put a resistor (500 ohm-1K) from the output cap on the speaker side to ground, and then a 5 second delay relay to the speaker.
The resistor stabilizes the "mid voltage" B+ charge across the cap allowing it to "charge", before the relay connects the speaker, resulting in no turn-on thumps.
I'm glad you said that. I was beginning to suspect our young friend had mistaken big power supply caps for output caps. That is likely the case.
Very plausible. If he asked if the amps were dc coupled, he may have been told they weren't and that gave him the idea the outputs were capacitor coupled.
Not really.
Regardless of the amp's design, it's always got two signal lines and a common signal ground.